Dual Role of Matrix Metalloproteinases in Brain Injury and Neurorepair after Cerebral Ischemia Feifei Ma DOCTORAL THESIS Barcelona Dual Role of Matrix Metalloproteinases in Brain Injury and Neurorepair after Cerebral Ischemia This thesis is submitted by Feifei Ma to apply for her doctoral degree at Universitat Autònoma de Barcelona. This work has been directed by Dr. Anna Rosell and Dr. Joan Montaner from Vall d’Hebron Institut de Recerca Directors: Doctoral student: Dr. Anna Rosell Novel Feifei Ma Dr. Joan Montaner Villalonga Barcelona, 2015 This study was conducted with funding from the following grants:  The Fondo de Investigaciones Sanitarias (PI13/01094)  The Spanish stroke research network INVICTUS (RD12/0014/0005) of the Instituto de Salud Carlos III  ERANET-NEURON program (PRPPRI-PIMN EU-2011-1352) from the Spanish Ministerio de Economía y Competitividad.  A pre-doctoral fellowship from the China Scholarship Council (CSC) Acknowledgment Acknowledgement Three years are really short for me to complete my PhD program and obtain many good achievements. But during these years, I have got a lot: the precious experience in a well-controlled laboratory, the specific analytical capability in the scientific study, the logical reasoning skill in solving problems, the good habits in research and daily life… I wish here to thank all members who have considerably supported me during these three years, particularly to my lab mates, my family and friends. Without any of you, I would not have finished my work today. To my supervisor Dr. Joan Montaner, for giving me this wonderful opportunity to do my PhD studies in his lab, an excellent place for working. Under his professional guidance, I have broadened my horizon in the area. To my direct thesis supervisor Dr. Anna Rosell, for all the supports I have received. She supervised me for the directions of my research, gave my valuable suggestions and comments of my study, taught me how to write a scientific paper and helped me to solve kinds of problems regarding to my study as well as my life. I couldn’t imagine how would I be today without her patience. I would like to give my gratefulness to Anna Morancho who is one of the most humorous and lovely girls I have met. She has contributed a lot for this work and I have learnt a lot from her. Thanks to Dolors, I am not getting crazy with those complicate statistical analyses; thank Marina, Vero, Pablo and people from the hospital, for their contributions in this project; thank other members in the lab for giving me a lot of help. To my parents and my brother, I have own them too much. Thanks for their supports and tolerances allowing me far away from them to pursue my dream. To my boyfriend, Wei, thanks him for every moment we have. Molt Gracies! Abstract Abstract Ischemic stroke is a leading cause of mortality, long-time disability and morbidity in the world. Currently, the only effective treatments are the intravenous thrombolytic therapy with tissue plasminogen activator (tPA) and the new emerging intra-arterial thrombectomies, which require to be administered in the hyperacute phase of stroke (<4.5 hours for tPA treatment and <6 hours for endovascular therapy). Therefore, it is necessary to investigate new therapies that could be used to treat a large number of patients in a later phase to repair and rewire the damaged tissue. One approach to achieve these objectives is to enhance vascular remodeling and neurogenesis in the ischemic brain. The aim of this thesis is to study the modulation of extracellular matrix metalloproteinase-13 (MMP-13) in brain injury and neurorepair and to investigate several MMPs as serving biomarkers to monitor motor and functional improvement during neurorehabilitation therapy. Using a distal middle cerebral artery occlusion model of ischemia in mice, we found that the modulation of MMP-13 is involved in ischemic damage and repair mechanisms. After ischemia-reperfusion, MMP-13 deficiency reduces infarct size, improves functional outcome and protects from hemorrhagic transformation, but impairs cortical tissue recovery by reducing proliferating neuroblast migration and vessel remodelling in a microenvironment with reduced trophic factors. At the same time MMP-13-silenced endothelial progenitor cells present an aberrant in vitro tubulogenesis function. As MMP-13 deficiency protects brain injury acutely and participates in post-stroke tissue recovery, this thesis also studies the temporal profile of plasma MMP-3, MMP-12 and MMP-13 during rehabilitation therapy to investigate their potential role as biomarkers of tissue recovery. Our results show that high levels of plasma MMP-13, as well as MMP-12, are strongly associated with stroke severity acutely, but not MMP-3. Interestingly, those patients with a better motor outcome during rehabilitation present elevated MMP-3 and decreased MMP-12/-13 levels. MMP-3 is closely associated with the better post-stroke motor recovery, indicating the potential as a biomarker of final functional improvement. These are important results that envision the use of biomarkers that could help to adjust, personalize and modulate neurorehabilitation therapies according to patients’ needs. In conclusion this thesis shows the role of MMPs in tissue injury and repair after stroke and the importance to monitor and modulate these proteases in favor of patients’ recovery. Resum Resum L'ictus isquèmic és una de les principals causes de mortalitat, discapacitat de llarg termini i de morbiditat al món. Actualment, els únics tractaments eficaços són el tractament trombolític intravenós amb activador del plasminogen tissular (tPA) i les novedoses trombectomies intra-arterials, que requereixen ser administrades en la fase hiperaguda de la malaltia (<4,5 hores per al tractament amb tPA i <6 hores per les emergents la teràpies endovasculars). Per tant, és necessari investigar noves teràpies que podrien ser utilitzades per tractar un gran nombre de pacients en una fase posterior de la malaltia per reparar el teixit danyat. Alguns enfocaments terapèutics apunten que per assolir aquest objectiu és necessari potenciar la remodelació vascular i la neurogènesi en el cervell isquèmic. L'objectiu d'aquesta tesi és l'estudi de la modulació de la metaloproteinasa de matriu-13 (MMP-13) en la neurorreparación i el d’investigar diverses MMPs com a biomarcadors que serveixin per monitoritzar la millora de la funció motora durant la teràpia neuro-rehabilitadora. Utilitzant un model d’isquèmia cerebral per oclusió de l’arteria cerebral mitja en ratolins, es va trobar que expressió de la MMP-13 està implicada en el dany i en mecanismes de reparació post-isquèmia. Després de la isquèmia-reperfusió, la deficiència en MMP-13 redueix la mida de l'infart, millora l’estatus funcional en la fase més aguda i protegeix de l’aparició de transformacions hemorràgiques, però interfereix negativament en la recuperació del teixit cortical mitjançant la reducció en la migració neuroblast en proliferació, en el remodelat vascular i en un microambient amb factors tròfics reduïts. Al mateix temps les cèl·lules endotelials progenitores amb un silenciament transitori de l’expressió de MMP-13 presenten una funció aberrant en assajos de tubulogènesis, in vitro. Com que la deficiència de MMP-13 protegia del dany cerebral en el model animal d’isquèmia al temps que participava en la recuperació del teixit en fases de reparació, aquesta tesi també estudia el perfil temporal dels nivells plasmàtics de MMP-3, MMP-12 i MMP-13 durant la teràpia neuro-rehabilitadora per investigar el seu paper com a potencials biomarcadors del teixit en recuperació. Els nostres resultats mostren que alts nivells de plasma MMP-13, així com MMP-12, estan fortament associats amb la severitat i extensió de la isquèmia, però no els de MMP-3. Curiosament, els pacients amb un millor status en la funció motora durant la rehabilitació presenten els nivells més elevats de MMP-3 i reduïts de MMP-12 / -13. També, els valors de MMP-3 estan estretament associats amb la millor recuperació motora després de l'accident cerebrovascular, el que indica el seu potencial com a biomarcador de la millora en la funció motora. Aquests resultats són importants per imaginar el futur ús d'un biomarcador per tal de poder ajustar, personalitzar i modular teràpies de neuro-rehabilitació d'acord a les necessitats dels pacients. En conclusió aquesta tesi mostra el paper de les MMPs en el dany i reparació de teixits després de l'ictus isquèmic i ressalta la importància de monitoritzar i modular aquestes proteases en favor de la recuperació dels pacients. Index Index Abbreviations ................................................................................................. i 1. Introduction ......................................................................................................1 1.1 Ischemic Stroke. ..................................................................................................... 1 1.1.1 Ischemic Cascade and Neuroinflammation..................................................... 2 1.1.2 Ischemia Reperfusion Damage........................................................................ 5 1.1.3 Animal Models for Ischemic Stroke ................................................................ 7 1.1.4 The Neurovascular Unit................................................................................... 8 1.1.5 Angio-neurogenesis .................................................................................... 10 1.2 Matrix Metalloproteinases ................................................................................... 14 1.3 Treatments for Ischemic beyond t-PA.................................................................. 16 1.3.1 Rehabilitation after Stroke .......................................................................... 17 1.3.2 Rehabilitation and Tissue Repair................................................................... 18 1.3.1 Cell-based Therapy ........................................................................................ 20 2. Objectives ........................................................................................................23 3. Methods and Results ......................................................................................25 Article 1 ...................................................................................................................... 27 Endothelial progenitor cells and revascularization following stroke. Article 2 ...................................................................................................................... 39 Matrix Metalloproteinase-13 Controls Neuroprotection and Neurorepair after Cerebral Cortical Ischemia in Mice. Article 3 ...................................................................................................................... 65 Plasma matrix metalloproteinase levels in stroke patients during intensive rehabilitation therapy. 4. Discussion ........................................................................................................87 4.1 MMP-13 deficiency protects mice from ischemia-reperfusion injury whereas high level of plasma MMP-12 and -13 is associated with stroke severity in ischemic stroke patients....................................................................................................................... 87 4.2 MMP-13 deficiency inhibits spontaneous neurorepair by reducing the number of proliferating neuroblasts and angiogenesis in peri-infarct areas, while decreased level of plasma MMP-12 and -13 is associated with better motor recovery in patients. .................................................................................................................................... 92 4.3 MMP-13 gene silencing impairs the angio-vasculogenic function of Endothelial Progenitor Cells which shape simpler and less connected vascular networks in vitro. .................................................................................................................................... 95 4.4 Elevated plasma MMP-3 is associated to motor and functional improvement during rehabilitation therapy after stroke. . .............................................................. 97 4.5 Future perspectives . ............................................................................................ 99 5. Conclusions....................................................................................................101 6. References .....................................................................................................103 Supplementary Information 1 ..........................................................................121 Curriculum vitae of the author Supplementary Information 2 ..........................................................................125 Additional publication and poster Abbreviations Abbreviations ADAMTS ADAmalysin-like Metalloproteinase with ThromboSpondin motifs Ang Angiopoietin BBB Blood Brain Barrier BI Barthel Index CAHAI Chedoke Arm and Hand Activity Inventory CNS Central Nervous System CXCL12 C-X-C motif chemokine 12 CXCR-4 C-X-C chemokine receptor type 4 DNA DesoxyriboNucleic Acid DWI Diffusion-Weighted Imaging EPCs Endothelial Progenitor Cells FAC Functional Ambulation Categories FGF Fibroblast Growth Factor FeCl3 Ferric Chloride FMA Fugl-Meyer Assessment GABA γ-Aminobutyric acid HT Hemorrhagic Transformation HI Hemorrhagic Infarctions IL InterLeukin KO Knock-Out MCA Middle Cerebral Artery MCAo Middle Cerebral Artery occlusion MMP Matrix MetalloProteinase MRC Medical Research Council MRI Magnetic Resonance Imaging NIHSS the National Institutes of Health Stroke Scale NO Nitrogen Oxygen OECs Outgrowth Endothelial Cells PDGF Platelet-Derived Growth Factor PH Parenchymal Hematoma PWI Perfusion-Weighted Imaging rCBF residual Cerebral Blood Flow RNS Reactive Nitrogen Species ROS Reactive Oxygen Species SDF-1 Stromal Derived Factor-1 i SS-QOL Stroke Specific Quality of Life SVZ SubVentricular Zone SGZ SubGranular Zone TIA Transient Ischemia Attack TIMP Tissue Inhibitors of MetalloProteinase t-PA tissue Plasminogen Activator TNF-alpha Tumor Necrosis Factor-alpha TTC 2,3,5-TriphenylTetrazolium Chloride VEGF Vascular Endothelial Growth Factor WHO World Health Organization 8-OHdG 8-HydrOxy-2’-deoxyGuanosine ii Introduction 1. Introduction 1.1 Ischemic Stroke Stroke is one of the leading causes of death and long-term disabilities worldwide. Every year, according to the World Health Organization (WHO), 15 million people suffer a stroke: nearly 5 million die and another 5 million are permanently disabled (www.who.org). Stroke occurs due to alterations with the blood supply to the brain: either the blood supply is blocked or a blood vessel within the brain ruptures. As a consequence of the sudden reduction of the physiological levels of blood flow to the brain, it results in a loss of oxygen and glucose followed by cell death, brain injury and the loss of neurological functions. This pathology is currently one of the most important health problems because its annual incidence is about 200 cases per 100,000 inhabitants and is the leading cause of death in women and second in men in Spain1, in addition to the major cause of long-term disability due to motor and sensory limitations. The main risk factors associated with stroke are hypertension, diabetes, atrial fibrillation, smoking and high cholesterol levels, being more prevalent after the age of 652. There are two types of stroke: ischemic strokes and hemorrhagic strokes (Figure 1), plus the transient ischemic attacks (TIAs) which last a relatively short time and usually causes no permanent injury to the brain (www.strokeassociation.org). In numbers, around 80% of all strokes belong to the ischemic subtype and the other 20% are caused by a hemorrhage. During an ischemic stroke, the blood supply to the brain decreases dramatically which usually happens when an artery is blocked by a blood clot or by the narrowing of the arteries. The part of brain that depends on the blood supply of the affected artery cannot sustain the normal cell metabolism due to the limited amount of oxygen and glucose, and the so called ischemic cascade is activated. 1 Figure 1. Ischemic and hemorrhagic stroke in human brain. The figure is adapted from Heart & Stroke Foundation (www.heartandstroke.com) In a hemorrhagic stroke, an artery in the brain bursts: blood vessels inside the brain leak causing an intra-cerebral hemorrhage, while blooding under the outer meningeal coverings of the brain and into the fluid-filled space surrounding the brain leading to a subarachnoid hemorrhage. The TIA, also called as “mini-stroke", is often characterized as a “warning stroke” as a stroke finally occurs with most of those patients presenting a TIA symptoms in the following hours to months3. The reduction of blood flow in TIA is temporary which might not cause real cellular death and clinical symptoms usually last for a few minutes. However, patients that have suffered a TIA have more vascular events than those who did not fulfill the criteria of TIA at either 30 days or 90 days after the initial events4. The work presented in this thesis focuses on human ischemic stroke and an experimental model of this disease, therefore the remaining introduction text on mechanisms, treatments and molecular pathways of injury and repair will focus on this type of stroke. 1.1.1 Ischemic Cascade and Neuroinflammation In ischemic stroke, the loss of blood supply into the brain results in a rapid loss of oxygen and glucose for the tissue, followed by the reduced cell viability or cell death accompanied by a destructured extracellular matrix that limits cell-cell interactions and the normal function of neuronal networks. As described below, all the process occurs in a cascade of pathological events: once the residual cerebral blood flow (rCBF) decreases below the 30% of the baseline 2 Introduction blood flow, neuronal dysfunction occurs. As the electrical function of the brain is intimately related to metabolism, a biochemical cascade is initiated, involving excitotoxicity, oxidative stress, inflammation and apoptosis5,6. Physiologically, the brain has a very high consumption of oxygen and glucose but little capability to store energy; hence the mitochondria play a key role during the metabolism in the brain. Once ischemia occurs, insufficient energy leads to membrane evoked potential cessation and cells depolarization, activating voltage/acid-sensing ion channels, leading to an imbalance in sodium and calcium levels (Figure 2). Calcium is thought to initiate a series of cytoplasmatic and nuclear events including targeting the intrinsic mechanical apoptosis7. Activation of calpains by increased calcium results in the cleavage of cell death related pro-apoptotic proteins. Actually, cells in ischemic core undergo necrosis and apoptosis: the initiation in ischemic core is thought to occur by an apoptotic pathway, but the severe decrease in energy level causes a shift from apoptosis toward secondary necrosis. At the same time, the energy-dependent processes such as presynaptic re-uptake of excitatory amino acids are impeded, which increase the accumulation of glutamate in the extracellular space and subsequent excitotoxicity. Induced cell damage occurs through necrotic and apoptotic cell death pathway after the triggering of glutamate receptors7. In parallel, mitochondria oxidative phosphorylation is also impaired due to the low levels of oxygen and glucose during ischemia, leading to the release of reactive oxygen radicals (or reactive oxygen species, ROS) and reactive nitrogen species (RNS) as well as increased anaerobic glycolysis6. ROS are considered as the main cause of membrane and fundamental cellular components damage8; and RNS act together with ROS forming highly reactive products which are capable of damaging lipids, proteins and DNA of the cells9. These direct damages on cellular membrane and other compartments, together with the indirect injuries from free radicals stimulating various cellular molecular pathways, lead to an abnormal accumulation of fluid within the brain parenchyma, which is called brain edema10. In addition, the overproduction of lactic acid from anaerobic glycolysis of the brain tissue causes neuron-toxic lactate accumulation, increasing the acute excitotoxicity phenomena which is also related to neuronal cell death11. 3 Figure 2. Summary of pathological mechanisms in the ischemic brain. The figure is adapted from Dirnagl et al. 1999. However, the immediate permanent cell damage/death after the severe reduction of blood flow occurs mainly within the center of the ischemic territory, which is called the “ischemic core”. The peripheral hypoperfused tissue, where the blood flow is inadequate to support neuronal function, but just adequate to maintain cell viability, might be salvageable and is called “ischemic penumbra”12. If functional hypoperfusion is not confirmed by neuroimaging techniques, the territory surrounding the infarct core should be named peri-infarct boundary/tissue. With time going on, without effective vessel reperfusion, the ischemic threshold increases and the infarct core expands into the penumbra which might be accompanied by neurological deterioration of the patient. Brain cells, particularly the vulnerable neurons, undergo apoptosis which dominates cell death during this expansion13. As shown in the Figure 3, the initial ischemic infarct lesion expands to the ischemic hypoperfused penumbra leading an extensive damage after several hours. Timely effective vessel perfusion or potential neuroprotective interventions can reduce the spreading of depolarization and decrease the final infarct size7,14. The penumbra therefore is still the target for any hyperacute stroke therapy, and to identify the penumbra where the tissue is at risk and to apply in-time treatments would be of great importance15. In accordance with this dynamics of infarct expansion, the only effective treatments to avoid infarct expansion are the thrombolytic therapies with recombinant tissue plasminogen activator (t-PA) that should be administered within the first few hours16, and the new endovascular treatments which are also required to be performed within the first hours after stroke onset17,18,19. 4 Introduction Figure3. Magnetic resonance imaging (MRI) from an ischemic stroke patient. Perfusion-weighted imaging (PWI) indicating the large area of hypoperfusion at <6 hours after stroke onset (A). Diffusion-weighted imaging (DWI) showing the ischemic core at the same level of PWI at <6 hours after stroke onset (B) and at 48 hours (C). Finally, other mechanisms are also involved in the process of ischemic stroke at later time-points, such as inflammation. The calcium activated second-messenger systems, the increased free radicals as well as hypoxia itself trigger the expression of numbers of pro-inflammatory genes and activation of these molecules7. A rapid activation of resident inflammatory cells followed by the infiltration of circulating inflammatory cells in ischemic brain region has been documented20. Besides, the presence of necrotic cells in the tissue is also highly immunostimulatory. As a consequence, a large amount of inflammatory factors including cytokines, chemokines, matrix metalloproteinases (MMPs), adhesion molecules and ROS are released from injured tissue. Increased evidences indicate that inflammation contribute to ischemic brain damage and predominantly is deleterious in the early phase after ischemic stroke20; but growing evidence suggests that several neurological repair mechanisms might be coupled to the neuro-inflammation stimuli21,22. Major steps in the repair process include removal of dead cells, development of an anti-inflammatory micro environment, and generation of pro-survival factors fostering tissue reconstruction and repair23,24. Very few knowledge regarding to the participation of inflammation in brain tissue recovery is established, and even less in the long-term effects. This thesis will explore the role of some matrix metalloproteinases and growth factors in the delayed phase of ischemic damage. 1.1.2 Ischemic Reperfusion Damage The reperfusion of vascular supply to an area temporarily deprived of blood flow is not always related to a final recovery of the tissue and good functional outcome, and 5 after stroke, there is usually some degree of reperfusion related to spontaneously reperfusion or treatment-induced recanalization of the occluded artery. Administration of t-PA 3 hours after clot-induced ischemia leads to effective reperfusion without affecting infarct volume or functional outcome at 24 hours after MCAo in mice25. Insufficient reperfusion or delayed recanalization is frequently associated with an exacerbation of tissue injury and profound tissue damage as explained in this section. During reperfusion after ischemia, both the innate and adaptive immune systems as well as the complement system, platelets and coagulation factors are involved26. Particularly during the early phase of reperfusion, innate immune cells dominate the cellular compositions of infiltrates such as resident microglia cells (the main innate immune cells in the brain), which response firstly to brain injury. Microglia exerting neurotoxic functions enhances the blood brain barrier (BBB) permeability by preceding the leukocyte infiltration. The BBB is constituted by vascular endothelial cells, basement membrane and pericytes, as well as the end-feet of astrocytes, by providing to the central nervous system a highly controlled cerebral homeostasis27. Both ischemia and reperfusion can destabilize and destroy the balance of this microenvironment by inducing the disruption of BBB integrity as proved in experimental models occluding the middle cerebral artery (MCA)28. But in transient ischemia models, the BBB disruption is typically more severe than that in permanent models, related to the sudden increased blood volume during reperfusion29. A major phenomenon during arterial reperfusion is the recirculation of oxygen which induces the overproduction of ROS and nitric oxide (NO) resulting in additional oxidative stress. On one hand cell death occurs through a number of mechanisms including both necrosis and apoptosis by mediating lipid peroxidation and desoxyribonucleic acid (DNA) damage induced by oxidative stress26. On the other hand impaired endothelial cell function occurs as a result of the excessive pericytes contraction and the exposure of the sub-endothelial extracellular matrix to blood flow, altering the permeability of BBB10. Subsequent hemorrhagic transformation (HT) of ischemic infarction is a well-recognized life-threatening consequence of ischemia followed by reperfusion. This phenomena occurs in around 10%-40% of ischemic stroke patients during ischemia-reperfusion, being the main complication of t-PA treatment30. Either early or delayed administration of t-PA cause different degrees of hemorrhage. From animal models of cerebral ischemia, the administration of t-PA 20 minutes after injury significantly increases the small HT within the infracted tissue (named hemorrhagic infarctions, HI), while 3 hours post-ischemia treatment increases the 6 Introduction occurrence of severe HT including type II HI and massive and expanding HT (named parenchymal hematomas, PH)25. In stroke patients, the presence of mild HI might not cause major neurological deterioration of the patient and it is considered a sign of benign reperfusion of the artery, whereas PH is usually associated with a worsening of the neurological status of the patient (especially those causing clinical symptoms in patients which represent 3-7 % of all HT events31). As mentioned, the disruption of the BBB and the continued endothelial damage mediated by free radicals contribute to the appearance of HT in ischemic-reperfusion tissue. However, the mechanism of early HT (within 18-24 hours after stroke onset) is different from the delayed HT (beyond 18-24 hours). The early HT is thought to relate to leukocytes-derived MMPs damaging the BBB, while the delayed HT associated with ischemia activated brain derived proteases, neuroinflammation and other factors30. Other consequences of the disruption of BBB are the aggravation of tissue swelling, capillary plugging, platelet adhesion and thrombosis, being those correlated with enhanced stroke damage by the communication within neurovascular units32. 1.1.3 Animal Models for Ischemic Stroke Animal models of cerebral ischemia are effective experimental tools to better understand the pathological mechanisms of ischemic injury and to test potential therapeutic interventions in preclinical trials. As different types of stroke occur in human associated to different mechanisms, etiologies and outcomes, there is no single universal animal model for all forms of human stroke33. There are multiple animal models of focal cerebral ischemia involving both permanent and transient middle cerebral artery occlusion (MCAo) mostly applied in rodents due to ethical and economic reasons. Intra-luminal filament models are well-established methods and widely considered adequate to reproduce primary ischemic insult and subsequent cell death, glial activation, blood brain barrier damage, among other pathological mechanisms34,35. But this model presents high premature mortality rates due to the large ischemic lesion affecting both the striatum and the cortex (Figure 4), produced by a proximal occlusion of the MCA, which might be not suitable for long-term neurorepair or safety studies. On the other hand, a different size of infarct lesion between young and aged mice is found only in distal MCAo models but not proximal, indicating that small infarct lesions are more willing to be modulated by neurogenesis-involved mechanisms36. Besides, many 7 human strokes are small in size and the occlusion of distal MCA is the most common reason for human ischemic stroke37,38. In experimental models, distal MCAo can be induced by permanent electrocoagulation, vessel contraction (endothelin injection) or by a transient ligation/compression of the distal branch of the MCA generating a cortical infarction. Among other described methods39, the embolic and thromboembolic models are ideal methods to mimic the clinical ischemic stroke, and offer the potential to evaluate thrombolytic agents or combination therapies40,41. However, these models induce occlusions without control of exact reperfusion time and degree. A focal cerebral ischemia model produced by ferric chloride (FeCl3)-induced thrombus formation has been described in a previous publication42. This model holds the advantage of low mortality, easy to perform and no additional trauma to the brain tissue. However, reperfusion is still hard to achieve and to control. Finally, a new reproducible method for distal focal cerebral ischemia in mice with reperfusion by the distal MCA compression was developed in our lab, based on the previous mechanical methods43. The present model holds the advantages of minimal damage to the vessel and underlying tissue, high reproducibility and low mortality, at the same time that allows controlled tissue reperfusion. Mortality is a critical parameter in experimental studies as low mortality rates are important to reduce the number of animals and avoid selective bias of the obtained results. This model together with the distal electro-coagulation of the MCA has been used as the experimental ischemia model in this thesis. Figure 4. 2,3,5-Triphenyltetrazolium chloride (TTC) stained brain sections non-viable tissue (white areas) after different experimental models of stroke in rodents. 1.1.4 The Neurovascular Unit Although the mechanisms of single cell death have been studied deeply, the mechanism by which the whole BBB responds to ischemia-reperfusion in the injury and recovery phases of stroke are still under investigation. The concept of neurovascular unit was suggested more than one decade ago, and refers to the 8 Introduction interconnection among cerebral endothelial cells, astrocytes, adjacent neurons and the extracellular matrix together with the circulating elements in the blood stream2. It is a functional unit revealing the complex inter-relationships among all the brain cells and tissue environment related to cerebral diseases. Neurons, as the core functional cell in the brain tissue, play a dominant role during the development of neurological dysfunction in cerebral ischemia; however, neurons are not the only target influencing tissue outcome in ischemia-reperfusion damage. The neurovascular unit emphasizes the dynamics communication of vascular, cellular and matrix signaling in the brain (Figure 5). Figure 5. Organization of the capillary neurovascular unit. The figure is adapted from Abbott et al. 2010. An example is the disruption of the BBB which is crucially dependent on endothelial-astrocyte-matrix interactions which are related to the tight junctions and basal lamina integrity. Endothelial cells, the main source of reactive radical species after ischemia-reperfusion, contribute to the BBB connection via adherent and tight junctions to form a physical barrier together with pericytes. Astrocytic end-feed wraps around capillary endothelial cells regulate cerebral capillary blood flow as well as the permeability. We now know that the endothelial cells and astrocytes engage in a complex cross talk to regulate each other’s function 44. Extracellular matrix proteins contribute to the structural formation of basal lamina that connects to endothelial cells by the cell-matrix interaction. When the BBB is disrupted by ischemia, injury to the microvasculature can lead to injuries to neuron in the same territory 45. In this context, ischemic brain damage is not the only trigger of neuronal cell death, also the local environment which influences surrounding cells, allowing neurons to survive or to die. The function of this neurovascular unit in the cerebral micro environment is also related to the “cross-talk” among cells46. Increased studies have documented that various factors mediate the interaction both in health and disease including neurotransmitters and inflammatory mediators47. For example, during synaptic 9 activities, being initiated by neurons, both neurons and astrocytes release neurotransmitters, including acetylcholine, glutamate and γ-Aminobutyric acid (GABA) directly or indirectly stimulating the production of vasodilators to control the vasodilatation of arteries at the site of activation44. But the homeostasis of the neurovascular unit is also important during stroke recovery. The expression of reactive astroglia expressed glutamate transporter is positive related with the neurological score during day 1-3 after ischemia-reperfusion in rats48. Post-stroke injection of antagonist of glia-specific glutamate transporter significantly changes the number of neurons responding to forelimb stimulation after stroke in mice49. 1.1.5 Angio-neurogenesis In spite of the neurological dysfunctions caused by stroke are a consequence of tissue and cell damage, surviving patients undergo some spontaneous recovery by triggering self-repair processes (Figure 6)50,51. However these are limited in terms of tissue repair. Recovery could therefore occur by multiple mechanisms52,53: 1) Restoration of damaged neurons and their function; 2) Remodeling the structure or function of the remained undamaged neural cells (so called “neuroplasticity”); 3) Formation of new functional neurons from pre-existing progenitors. Figure 6. Representation of injury and spontaneous repair mechanisms triggered by stroke. The figure is adapted from E Lo et al, 2008. 10 Introduction The formation of new functional neurons is called neurogenesis. Physiologically, newborn neurons generated from the subventricular zone (SVZ) migrate to olfactory regions whereas those neurons generated from subgranular zone (SGZ) migrate to hippocampus54. It has been reported that after cerebral ischemia, SVZ-derived neural precursor cells have the capability to respond to the insult and migrate toward to the lesion area to produce new cells replacing the damaged cells in rodents55,56, and this migration and proliferation was also observed in human in response to ischemic stroke57. This generation of newborn neurons after stroke is proven associated with functional recovery by integrating into the existed neuronal network58. However, the sustained functional improvement is obviously earlier than the maturation of these new generated neurons 59. Besides, the majority of these newborn cells die from apoptosis in a few days, and the rate of differentiation is low55,60,61. This mismatch supports the hypothesis that neurogenesis induced-recovery is mediated by mechanisms other than directly neuronal replacement, possibly through local growth factors mediated functional restoration in the peri-infarct tissue59. As detailed in previous sections, ischemia initiates inflammation, increases microvascular permeability and might cause hemorrhage, which does not only influence neurons in the first stage but also other glial cells supporting neurons and the microvessels that supply oxygen and glucose. The concept of the neurovascular unit proposes that, injury to any part of the unit affects the other components unbalancing the neurovascular homeostasis that allows the physiological environment of the brain. In this regard, the most classical view of neurorepair as the replacement of dying neurons has changed in the last decades for a more global understanding of the brain, including other cell types (glial cells, inflammatory cells and stem/progenitor cells) but also the extracellular matrix and their interactions. Therefore endogenous mechanisms of neurorepair include angio-vasculogenesis (formation of new blood vessels), gliogenesis (formation of new glial cells), neurogenesis (formation of new neurons), re-myelination (new myelin sheaths on demyelinated axons), among others. Part of this thesis focuses on angiogenesis as a key mechanism to modulate during stroke repair. Classically, the formation of new blood vessels was thought to be mediated exclusively by embryogenic vasculogenesis, followed by the sprouting of endothelial cells from pre-existing vessels during angiogenesis62. However, this dogma was overturned with the discovery of bone marrow-derived CD34+ cells with endothelial characteristics and circulating in adult human blood63. These cells, referred as 11 endothelial progenitor cells (EPCs), are capable of differentiating ex vivo into endothelial-phenotyped cells, and represent a new model for endothelial regeneration and vessel repair. In this context, EPCs have been shown to be highly migratory to ischemic areas and a powerful therapeutic potential by directly enhancing angiogenesis or by forming new vessels through vasculogenesis64,65. In addition to the therapeutic interest of endothelial progenitors for vascular regeneration, it is extremely important to know more on the role played by the growth factors secreted by these cells which enrich the so called neurovascular niche. It is known that angiogenesis-related genes such as vascular endothelial growth factor (VEGF) are up-regulated as soon as 1 hour after ischemia onset; and the post-stroke vascular formation in the peri-infarct area is also involved in the process of tissue recovery, involving angiopoietins (Ang) up to 21 days after ischemia60,66,67. But angiogenesis activation after cerebral ischemia does not occur alone, it is actually coupled to neuroblast cell migration from the SVZ58. In this context, the role might be as important as the one played by neural precursor cells. The so called neurovascular niche defines these complex mechanisms of crosstalk between cerebral endothelium and neural precursors under ongoing neurogenesis. Stroke changes the cellular environment in the peri-infarct area by forming a unique neurovascular niche, in which angiogenesis and neurogenesis are linked through specific vascular factors and chemokins59. Apart from the VEGF, the cytokine stromal derived factor-1 (SDF-1), also named as C-X-C motif chemokine 12 (CXCL12), is another important molecular playing a central role in the post-stroke neurovascular niche. The SDF-1 and its receptor C-X-C motif chemokine receptor type 4 (CXCR-4) participate in a variety of physiological and pathological processes including angiogenesis and neurogenesis through mediating the proliferation of neurogenitors, regulating the migration, differentiation, as well as functional integration of newborn neurons into existing networks68. We have also learnt that cerebral vessels express SDF-1 days after cerebral ischemia in peri-infarct areas where neurogenesis occurs, and the specific blockage of angiogenesis by the administration of specific inhibitors (such as endostatin) also blocks endogenous post-stroke neurogenesis from the SVZ59. As we will see in this thesis, several therapeutic approaches under investigation have their target on the modulation and enhancement of these mechanisms that participate in neurological repair after stroke. 12 Introduction 1.2 Matrix Metalloproteinases The degradation of extracellular matrix substrates plays a crucial role in both ischemia induced disruptions of the BBB as well as in cell proliferation and cell migration during tissue repair. Together with adamalysin-like metalloproteinases with thrombospondin motifs (ADAMTSs), MMPs are capable to degrade the extracellular matrix, plating important roles in many physiopathological processes69,70. MMPs are a family of zinc-dependent proteinases, involved in the degradation and remodeling of extravascular matrix and vascular compartment due to their capability to degradation of basal lamina and extracellular matrix components. This protease expression and activation occurs in physiological conditions during the normal homeostasis of the extracellular matrix but the dysregulation of some MMPs also participates in processes of tissue injury and repair71. Despite MMPs have been considered for a long time to be the prototypic matrix-degrading proteases, in recent years and with the application of engineered genetic models, they have been shown functions more as processing enzymes that control numerous, diverse cell processes. Indeed, via cleavage of cytokines, chemokines, antimicrobial peptides, and cell surface proteins, MMP proteolysis affects critical regulatory functions related to immunity, tissue repair, differentiation, and cell transformation72. There are four primary subgroups of MMPs on the premise of the structure domain: collagenases (MMP-1, -8 and -13), gelatinases (MMP-2 and -9), stromelysins (MMP-3, -10 and -11) and others not belonging to any of them73. Physiologically, most of the MMPs are synthesized as inactive forms (also named pro-forms), and their activity is strictly controlled by gene transcription which is mediated by variety of inflammatory cytokines, hormones and growth factors74. However, MMPs are also regulated at post-transcriptional level through an activation mechanism: the activated MMPs can participate in processing the activation of other MMPs, and the irreversible MMPs’ activation can be blocked by tissue inhibitors of metalloproteinase (TIMP)75. Uncontrolled MMP activity has been described in several diseases and many pathological processes including cerebral ischemia are associated to an increase in MMP levels and activity related to the degradation of proteins such as laminin, fibronectin, collagen, proteoglycans, and others. Different investigators have suggested that the upregulation of MMPs such as MMP-2, MMP-3 and MMP-9 contribute to infarct extension and BBB breakdown after stroke76,77,78, and blood levels of some MMPs can also be easily monitored as potential biomarkers of brain tissue damage and/or neurological outcome79,80. Also, 13 brain MMP-9 is up-regulated in ischemic brain tissue related to vessel disruption and blood extravasation in human samples81,82,83, and the gene or pharmacologic inihibition of MMP-9 has demonstrated to protect the brain from cerebral ischemia in preclinical models84,85,86. A list of the putative role of most relevant MMPs related to ischemic stroke has already been summarized by some researchers (Table 1)87. Table 1. MMPs and their putative roles in ischemic stroke. (The table is adapted from Lakhan et al. 2013) MMP Putative roles in ischemic stroke MMP-2 Attacks major components of the basal lamina around the cerebral blood vessels (gelatinase A) including collagen IV, lamina and fibronectin. May contribute to infarction and hemorrhagic volume. MMP-3 Degrades the extracellular matrix proteins fibronectin, denatures collagen, (stromelysin-1) lamina and proteoglycans. Plays a key role in the initial opening of the BBB after stroke and the development of hemorrhagic transformation, particularly with tPA treatment. Promotes remyelination and the chemokines induced neural progenitor cells’ differentiation and migration. MMP-8 Preferentially cleaves collagen I. Unregulated in infracted tissue but studies are (collagenase-2) limited. MMP-9 Attacks major components of the basal lamina including collagen IV, lamina and (gelatinase B) fibronectin. Plays a key role in the delayed opening of the BBB especially in the states of systematic inflammation, related to hemorrhagic transformation particularly with tPA treatment. Is involved in the neurovascular remodeling by mobilizing VEGF, myelin reformation and EPCs angiogenic function; promotes the chemokines induced neural progenitor cells’ differentiation and migration. MMP-12 Degrades soluble and insoluble elastin. Contributes to the acute brain damage (macrophage may related to the activation of MMP-9. metalloelastase) MMP-13 Preferentially cleaves collagen II. Upregulated in infracted tissue but its study is (collagenase-3) limited. At the beginning of this section we have pointed out that these proteases might also contribute to beneficial remodeling during stroke recovery. In this regard, early studies already shown that MMPs are expressed during development and contribute 14 Introduction to morphogenesis of the CNS88. They are also related to neurogenesis mediating the differentiation and chemokine-induced cell migration of adult neural progenitor cells89. MMPs may also modulate bioavailable levels of various growth factors (such as VEGF by MMP-9 actions) by processing pro-form precursors or by liberating active molecules from matrix-hidden compartments90. After stroke, newly born neuroblasts migrate from the SVZ to peri-infarct cortex, and increased vascular remodeling is also found in this area91,92. At 2 weeks after cerebral ischemia in mice, MMP-9 is enhanced in the SVZ and co-localized with BrdU-labeled cells and neuroblasts whereas the pharmacological inhibition of MMPs reduces the extension of neuroblast signals that extend from the SVZ into the damaged striatum and post-stroke angiogenesis impairing functional recovery93,94. As we will better describe latter, our group has recently described the central role of MMP-9 in vascular remodeling mediated by endothelial progenitor cells in the context of cerebral ischemia both in vitro and in vivo95,96. Regarding MMP-13 (collagenase-3), the dominant collagenase, the current knowledge comes from cancer, bone-related diseases and re-epithelialazation of the skin during wound healing97,98,99, but limited studies have been published related to the role of this protease in stroke disease. A few studies from our group and others have shown an increased expression of MMP-13 in infarct areas compared to healthy control human tissue81, located in neurons as well as in glial cells in ischemic models100,101. Also a high level of MMP-13 in blood is related to the early lesion expansion in stroke patients who received thrombolytic therapy102. The modulation of MMP-13 during the delayed phases of stroke is still unknown. 15 1.3 Treatments for Ischemic beyond t-PA Although the understanding of post-ischemic injury is still limited, new insights of therapeutic approaches are accumulating. Being the blockage of a cerebral artery by a blood clot, the pathological cause of ischemic stroke, the recanalization therapy to reperfuse the artery and recover the cerebral blood flow is still the only effective treatment to reduce tissue injury and improve neurological outcome. The intravenous thrombolysis with t-PA is still the mainstay of acute ischemic stroke management for the patients within 4.5 hours from symptom onset, being a standard therapy for those patients with moderately severe symptoms younger than 80 years old and without contraindications for the medication16,103. The new emerging effective intra-arterial therapy using stent retrievers also requires to be applied within the first 6 hours17,103. As both of these two successful treatments require rapid intervention, only 5% to 7% of patients are treated with t-PA along with 1% to 2% benefit from intra-arterial therapy, whereas the majority of patients do not meet criteria to receive these available reperfusion therapies. In this context, the development of neurorestorative and neurorepair therapies available for the vast majority of patients to improve neurological deficits after ischemic stroke is a great challenge. As mentioned, the spontaneous brain neurorepair is limited in efficacy and heterogeneous among patients. Strategies to enhance the endogenous restorative mechanisms of recovery are sought as stroke treatments. The central goal of these potential therapies should be to prevent further damage during the delayed phases of ischemia minimizing the damaging molecular events and cell death, at the same time that new opportunities of tissue recovery should be stimulated to remodel and recover the affected tissue (Figure 6). Figure 6. Spectrum of therapies for ischemic stroke according to the chronical sequence of 16 Introduction mechanisms of injury and repair. The figure is adapted from Zaleska et al. 2009. 1.3.1 Neurorehabilitation after Stroke Despite important advances in acute stroke management in the last two decades (the management of patients in stroke units, the thrombolytic therapy or the new emerging endovascular treatments), each year about five million people survive a stroke worldwide with some disabilities that limit their independence for daily activities (www.who.int). This fact also carries an enormous socio-economic burden estimated at ‎€196 billion every year for the countries of the European Union (www.ehnheart.org). For those patients the only approved treatment during the sub-acute and chronic phases of the disease is neurorehabilitation to recover the impaired or lost neurological functions with the objective of improving their independency status and quality of life104,105. During rehabilitation, the improved outcome in patients is the result of early mobilization, rehabilitation intensity and other measures to prevent/treat medical and psychological complications, altogether being performed by an inter-disciplinary team, and also according to spontaneous individual and heterogeneous response to therapy106,107. And the ultimate goal of any rehabilitation therapy is to reach the maximum recovery of the impaired function in an effort to reduce disability and achieve functional independence for daily activities of the patient. However, the recovery of other non-motor functions (the executive function, attention, memory, etc.) is also the subject for rehabilitation therapies. Importantly, for the overall rehabilitation process it is very important to assess patient's neurological function, independence and quality of life after stroke. The impact of the disease can be measured using certain scales which also allow to monitor and to evaluate the impact of the received therapies. As there is no consensus on the use of a single scale that can measure all the affected aspects of disability after stroke, many scales are commonly used in the clinical practice and in clinical trials to evaluate the therapeutic potential of a particular therapy105,108 despite certain limitations in their sensitivity, specificity and reliability among observers. It is therefore common to use a battery of scales to assess neurological deterioration or improvement of the patient’s function in rehabilitation therapy104,109,110. Table 2 shows a summary of the most used scales during rehabilitation therapy after stroke. 17 Table 2. Summary of the most used tests/ scales to measure the outcome in stroke rehabilitation. Measures Description NIH Stroke Scale (NIHSS) A quantitative measure of the severity of symptoms/ neurological deficit associated with cerebral infarcts. Fugl-Meyer Assessment (FMA) A disease-specific impairment index to assess motor function, balance, sensation qualities and joint function. Medical Research Council (MRC) To exam the peripheral nerve lesions by testing the muscle strength. mRankin A global outcomes rating scale for daily activity/disability. Barthel Index (BI) The most widely used measure of functional disability. Functional Ambulation Categories To assess ambulation capability. (FAC) Chedoke Arm and Hand Activity A functional assessment for the upper-limb recovery. Inventory (CAHAI) 10 meters’ Walking A sub maximal test to assess functional capability. Stroke Specific Quality of Life Scale To assess health-related quality of life. (SS-QOL) 1.3.2 Neurorehabilitation and Tissue Repair From a biological perspective, we know that neurorehabilitation enhances different aspects of brain plasticity such as axonal remodeling, cell genesis, angiogenesis or synaptogenesis111, which could be patient-dependent, and the biological mechanisms responsible for these individual responses to treatment are still being investigated. Fortunately, in some patients the recovery of the neurological function occurs spontaneously after stroke and in others it is achieved with some types of rehabilitation therapy up to 6 months or 1 year after stroke109. The spontaneous recovery is related to brain plasticity and is also observed in animal models of stroke being particularly significant in mice112. In this sense, several mechanisms have been 18 Introduction described to be related to functional recovery such as: the activation, proliferation and migration of endothelial/neural progenitor cells, changes in existing neural pathways, the compensation phenomena and also an increased neuroanatomical plasticity leading to the formation of new neuronal connections111. Today we know that the brain repair/regeneration process occurs from outlying areas of the infarct core, where these stem/progenitor cells arrive to replace or to provide trophic support to the pre-existing cells traffic113. Another important aspect during rehabilitation is the convenience to have biomarkers that could support an accurate and individual prognosis of stroke recovery during the rehabilitation program, which would help to decide on the intensity, type or duration of the rehabilitation programs108. However, very few studies have investigated the use of biomarkers during rehabilitation therapy for post-stroke recovery linked the progress of functional rehabilitation (determined by specific scales). In this regard the levels of certain cytokines such as interleukins (IL), tumor necrosis factor alpha (TNF-alpha) and adhesion molecules have been monitored in patients with a first-ever stroke under rehabilitation therapy compared with a control population showing an increase in IL-6 and TNF-alpha in stroke patients versus controls, and a negative association of IL-1-alpha, IL-8 or TNF-alpha with the Barthel Index114. But the study only covers the first two weeks after stroke. More recently, Blicher and colleagues have studied the neurotransmitter GABA by magnetic resonance spectroscopy in relation to motor function after 2 weeks of constrained-induced movement therapy showing an association between motor improvement and the magnitude of the GABA/creatine ratio115. Finally a recent study has reported the use of 8-hydroxy-2’-deoxyguanosine (8-OHdG), a sensitive marker for DNA damage that is detectable in urine samples, in a cohort of stroke patients undergoing rehabilitation therapy reporting a negative correlation between FMA scores and 8-OHdG level after rehabilitation treatment and lower levels of baseline 8-OHdG in patients that exhibited more improvement in motor function, suggesting the use of 8-OHdG as a biomarker of stroke recovery116. Given the complexity of the CNS responses to cerebral ischemia and the role of neural plasticity in tissue regeneration and functional recovery, it is likely that future rehabilitation treatments will be combined with pharmacological/biological therapies to enhance certain molecular pathways. In this context, to identify the key molecules responsible for these tissue regeneration mechanisms will be of great importance to influence the natural history of the disease. Part of this thesis focuses on the study of matrix metalloproteinases during intensive rehabilitation therapy related to the 19 motor/functional improvement achieved at different moments of the rehabilitation program. 1.3.3 Cell-based Therapy Cell-based therapies have been demonstrated to be an effective approach for brain repair by enhancing the endogenous restorative mechanisms of recovery in preclinical models of stroke117, and some of them are being tested in the clinical trials (/www.clinicaltrials.gov). Most of these treatments do not aim to replace the damaged cells, but to remodel the total system by modifying the cerebral microenvironments, thereby enhancing the neuroplasticity through angiogenesis, neurogenesis, oligodendrogenesis or synaptogenesis. Multiple stem/progenitor cells have been tested as potential sources for cell-based therapies including neural stem cells, mesenchymal stem cells or bone marrow-derived cells proving their efficacy in animal models of stroke improving the functional recovery or reducing brain lesions118. However, there are several types of stem cells, and to identify which subtype is the best one contributing to the robust therapeutic improvement is still a challenge. Moreover, there is still no consensus regarding to issues related to the therapy such as how long will these cells survive, what time point is the best one to administer those treatments or which group of patients are expected to benefit most from these therapies119. As the main goal of cell-based therapy is to create a favorable microenvironment for tissue repair in the desired peri-infarct area, and angiogenesis plays a pivotal role during this process, EPCs have attracted increasing attention in the last years. EPCs were firstly identified as bone marrow derived cells responsible for the postnatal vasculogenesis both in physiological and pathological neovascularization, circulating in adult humans120,63. After ischemic damage, EPCs mobilize from bone marrow to the peripheral blood; home to the sites of vascular remodeling in the brain where they differentiate into mature endothelial cells and/or regulate preexisting endothelial cells via paracrine or juxtacrine signals121, contributing to neurovascular repair after stroke. Importantly, the number and function of EPCs could be modulated after stroke to improve cell-based tehrapies122. In this line our group has demonstrated that the genetic or pharmacological inhibition of MMP-9 diminished the pro-angiogenic actions of EPCs95, and that the post-stroke neurovascular niche requires MMP-9 for an appropriate vascular remodeling and pro-angiogenic action of an EPC-based therapy in a mouse model of stroke96. These studies highlight the role 20 Introduction of MMPs in post-stroke neurorepair cell-based therapies. 21 22 Objectives 2. Objectives The main objectives of this PhD thesis are: 1) To conduct a systematic review of the literature studying Endothelial Progenitor Cells role in vascular remodeling and cell-cell interactions related to neurorepair after stroke. 2) To determine the role of MMP-13 both in brain injury after cerebral ischemia and in post-stroke neurorepair mechanisms such as angiogenesis and neurogenesis. 3) To identify the temporal profile of MMP-3, MMP-12 and MMP-13 serving as biomarkers of recovery during intensive rehabilitation therapy after ischemic stroke. 23 24 Methods and results 3. Methods and Results 25 26 Methods and results Article 1 Endothelial progenitor cells and revascularization following stroke Brain Research. 1623 (2015) 150-159. doi: 10.1016/j.brainres.2015.02.010. 27 28 29 30 31 32 33 34 35 36 37 38 Methods and Results Article 2 Matrix Metalloproteinase-13 Controls Neuroprotection and Neurorepair after Cerebral Cortical Ischemia in Mice Neurobiology of Disease, under review. Manuscript No.: NBD-15-907 39 40 Matrix Metalloproteinase-13 Controls Neuroprotection and Neurorepair after Cerebral Ischemia in Mice Feifei Ma, Pablo Martínez-San Segundo, Verónica Barceló, Anna Morancho, Marina Gabriel-Salazar, Dolors Giralt, Joan Montaner, Anna Rosell Neurovascular Research Laboratory and Neuroscience Department, Vall d’Hebron Research Institute, Universitat Autònoma de Barcelona, Spain. Passeig Vall d’Hebron 119-129, 08035, Barcelona, Spain Corresponding Author: Anna Rosell, anna.rosell@vhir.org Neurovascular Research Laboratory, Vall d’Hebron Research Institute, Passeig Vall d’Hebron 119-129, 08035, Barcelona, Spain. Tel. +34 934894029 ABSTRACT peri-infarct area (p<0.01). This result occurred in parallel with aberrant cortical vascular remodeling: post-stroke peri-infarct vessel New neuroreparative and neuroprotective density increased in the WT mice (p<0.01) but therapies are being sought to treat stroke this increase was blocked in the MMP-13 KO patients. One approach is the remodeling of mice. Prior to these vascular alterations, the extracellular matrix, which participates in both levels of pro-angiogenic factors, including G-CSF, brain injury and neurovascular repair when VEGF-A and angiopoietin-2, were lower in the matrix metalloproteinases (MMPs) are thought ischemic cortex of MMP-13 KO mice than in WT to be key players. Our aim was to investigate the mice (p<0.05). In vitro, gene-silencing of MMP-13 role of MMP-13 (collagenase-3) in the acute (24 in endothelial progenitor cells (EPCs) confirmed hours and 3 days) and delayed (2 weeks) phases the reduced ability of these cells to build of stroke. Permanent and transient cerebral tubulogenic networks in MatrigelTM substrate. ischemia models involving the cortex were Together, our results indicate that MMP-13 is a induced in MMP-13 knock-out (KO) and wild-type central protease in infarct development and (WT) mice. In the transient model, MMP-13 cortical remodeling during post-stroke deficiency reduced the amount of TTC-stained neurorepair, which is critical for optimal infarct tissue, reduced hemorrhagic events and angiogenic and neurogenic responses. improved functional outcomes (p<0.01). At two weeks, normal neuroblast (DCX+) migration from KEY WORDS: cerebral ischemia, MMP-13, the subventricular zone toward the peri-infarct angiogenesis, neurorepair, neurogenesis area was observed. However, MMP-13 deficiency significantly reduced the number of newborn neuroblasts (DCX+/BrdU+) in the cortical 41 INTRODUCTION that trophic factors are secreted by endothelial cells (Ohab et al., 2006) and glial cells (Hayakawa Stroke is one of the most common causes of et al., 2014). death and disability worldwide. Our investigation focused on ischemic stroke, which occurs when In this context, extracellular matrix remodeling is blood flow into the brain is interrupted by an a key step in angiogenesis and neurogenesis artery occlusion. Recombinant tissue during which matrix degradation is required to plasminogen activator (rtPA) and mechanical allow cell migration and capillary reorganization thrombectomy interventions are the only (Bovetti et al., 2007; Senger and Davis, 2011). therapies that have shown efficacy, but they Matrix metalloproteinases (MMPs), a family of must be administered within a short therapeutic zinc-dependent enzymes, are involved in the time window (Hacke et al., 2008; Lees et al., 2010, regulation of the composition of the cell-matrix. Jovin et al., 2015). New post-stroke therapies are Interestingly, some MMPs have been described therefore needed to diminish damaging not only as the focus of many neuroprotective molecular events and cellular death, which are pre-clinical therapies aimed at protecting the responsible for further damage, at the same time brain during the acute phase of stroke but also as that new opportunities to recover tissue are playing multiphasic roles, including roles in investigated. Of special interest are treatments neurovascular remodeling during neurological that induce neurological repair with the aim of recovery in rodent models of stroke (Zhao et al., improving functional outcomes and rescuing the 2006; Rosell and Lo, 2008). Among all the MMPs, limited capability the damaged brain has to the role of MMP-13 (also named collagenase-3) self-recover, with the potential to be in neurorepair and vascular remodeling is administered in a wider therapeutic time window unknown. As a major collagenase, MMP-13 (Zhang and Chopp, 2009). One approach to mainly cleaves collagen II, shows the highest potentiating neurological repair is to target gelatinase activity among the collagenases angioneurogenesis, which is the formation of (Knäuper et al., 1996; 1997) and has a central new vessels and neural cells. Adult neurogenesis position in the activation of other MMPs, occurs primarily in the subventricular zone (SVZ) through which it plays a key role in angiogenesis and the subgranular zone (SGZ), from which and the re-epithelialization of the skin during newly born neural cells migrate to the olfactory wound healing (Hattori et al., 2009). In the bulb or to the dentate gyrus under physiological context of cerebral ischemia, increased conditions (Ming and Song, 2011). However, expression of MMP-13 has been described in rat ischemic stroke alters this pattern of neurons, where it peaks on day 7 after ischemia neurogenesis and stimulates neuronal cells to (Nagel et al., 2005). Previous results from our migrate toward the damaged area (Ohab and group showed that high blood levels of MMP-13 Carmichael, 2008). In this environment, the were associated with lesion expansion in the neurovascular niche has been described as acute phase of stroke (Rosell et al., 2005), linking angiogenesis and neurogenesis, whereas nuclear MMP-13 was found to be demonstrating that, after stroke, neurogenesis expressed in neurons and glial supporting cells in occurs following angiogenesis in the peri-infarct human and rodent ischemic brain tissue areas and showing strong support for the idea (Cuadrado et al., 2009a). In the present study, we 42 aimed to determine whether MMP-13 could period. All procedures were approved by the acutely modulate neuroprotective and Ethics Committee of Animal Experimentation neurological repair functions during the recovery (CCEA 67/11) of the Vall d’Hebron Research phase of ischemia, and we focused our interest Institute and were conducted in compliance with on post-stroke vascular remodeling. For this Spanish legislation and in accordance with the purpose, both permanent and transient ischemic Directives of the European Union. stroke models were induced in MMP-13 deficient mice, demonstrating that MMP-13 is involved in Genotyping analysis acute infarct expansion at the same time that it is Freshly obtained mouse tail tissue was used for required for neurorepair during recovery. To DNA purification 116 to confirm the genotypes of further study the role of MMP-13 in vascular randomly selected mice in our colonies. Genomic remodeling, MMP-13 mRNA expression was DNA was collected and purified using a Gentra transiently silenced in outgrowing populations of Puregene Mouse Tail Kit (Qiagen, Germany) endothelial progenitor cells (EPCs), confirming according to a protocol provided by the the role of this collagenase in tubulogenesis manufacturer. Polymerase chain reaction was assays. Our data suggest that MMP-13 is a key performed using Platinum® Taq DNA polymerase protease that can be therapeutically modulated (Life Technologies, USA). The following pair of during both the acute and the recovery phases of Exon primers were used to genotype the stroke. wild-type mice: (Exon 5F) 5´TTTATTGTTGCTGCCCATGAG3´ and (Exon 6R) 5´AGTTTCTCCTCGGAGACTGGT3´. For the MATERIALS AND METHODS MMP-13 KO mice, a primer for the neomycin resistance gene, (neo) Animals 5´GACCCACCCCTTCCCAGCCTCT3´, was used as Age-matched 8- to 15-week-old male MMP-13 the oligonucleotide primer. The approximate knockout mice (MMP-13 KO in a C57BL/6 fragment sizes of each band were 1300 Kb for the background) and wild-type (WT, C57BL/6) were wild-type and 1485 Kb for the MMP-13 KO on bred in-house in a temperature and humidity agarose gels. controlled room and maintained on a 12 hour Focal cerebral ischemia models (MCAo) light/dark cycle. The offspring were used for these experiments. The founder mice for our WT The mice were given free access to food and and KO colonies were the kind gifts of Dr. water before surgery. Permanent Lopez-Otin and were generated as previously electrocoagulation and transient compression of described (Inada et al., 2004) and crossed into the distal part of the middle cerebral artery (MCA) congenic C57BL/6 mice for several generations. were performed as previously described by our Genotyping analyses were performed to confirm group (Morancho et al., 2012). Briefly, the that the mice were MMP-13 null, as described animals were anesthetized using 5% isoflurane below. A total of 99 animals (49 WT and 50 followed by 1.5 - 2% isoflurane for maintenance MMP-13 KO) were used in the present study, and anesthesia via a facemask in medical air (79% no mortality was observed during the study N2/21% O2; Abbott Laboratories, Madrid, Spain). 43 A small craniotomy was performed in the left CBF to below 80% and subsequent reperfusion to temporal bone (between the retro-orbital and above 75% of the baseline CBF were included in ear area) to expose the MCA. This artery was this study. Buprenorphine (0.05-0.1 mg/kg), an compressed for 60 minutes with a 30-G needle analgesic, was administered subcutaneously using a micromanipulator to perform transient immediately after the procedure. The animals in MCA occlusion (tMCAo) or electrocauterized for the sham group (n=6 for each genotype) were permanent MCA occlusion (pMCAo). A heating subjected to the same procedures (including 60 pad connected to a rectal probe was used to minutes of anesthesia for the tMCAo mice), maintain body temperature between 36.5-37.5°C except for electrocauterization or compression of during all surgical procedures. Cerebral blood the MCA. flow (CBF) was monitored using laser-Doppler flowmetry (Moor Instruments, UK) to ensure A total of 17 mice were subjected to pMCAo (8 appropriate occlusion in both models and WT and 9 MMP-13 KO), 70 to tMCAo (35 WT and reperfusion in the tMCAo mice (Morancho et al., 35 MMP-13 KO) and 12 to sham surgery (6 WT 2012). Only the mice that showed a reduction in and 6 MMP-13 KO), according to the design of the study, as presented in Figure 1. Figure 1. The study design. Three experimental protocols (A, B and C) were followed to study the role of MMP-13 in neuroprotection and neurological repairs after cerebral ischemia. Grip strength neurological test force used at one time point. This test was conducted blindly by a trained researcher before The grip strength test is designed to assess the the surgery (pre-surgery) and then repeated at peak forelimb force executed by a mouse when 24 hours, 3days, 7 days and 14 days the mouse releases its front paws from a grid. A post-ischemia (according to the study groups, A computerized grip strength meter (Harvard and B, as detailed in Figure 1). Apparatus, USA) was used. The specific protocol is described elsewhere (Rosell et al., 2013b). A Infarct volume and cortical atrophy total of 6 trials were conducted for each test, and the mean value was calculated as the average The mice that were subjected to 60 minutes of 44 occlusion (tMCAo) and the mice in the pMCAo Immunohistochemistry (IHC) group that were included in the infarct volume The ischemic animals in protocol B received protocol (study A in Figure 1) were transcardially 5-Bromo-2’-deoxyuridine (BrdU, 50 mg/kg in perfused with ice-cold saline and euthanized at saline, B9285, Sigma-Aldrich, USA) 24 hours or 3 days after ischemia. The dissected intraperitoneally every 2 days after ischemia. On brains were cut into 1 mm-thick coronal sections day 7 or day 14, the animals were euthanized, and then stained with 2.5% and the brain sections were prepared for 2,3,5-triphenyl-2H-tetrazolium chloride (TTC; immunohistochemistry as described for the Sigma-Aldrich, USA) for 15 minutes at room atrophy study. Later, the sections were fixed in temperature. The slices were post-fixed in ice-cold acetone and then blocked in 0.1% paraformaldehyde (PFA, 4%) and digitalized to PBS-tween containing 1% BSA and 10% goat obtain the measurements. The extension of the serum after acid denaturation with 2 M HCl PBS pale colored ischemic lesion was calculated for 1 hour. Slides were incubated with the blindly, and the results are presented as the following primary antibodies overnight at 4°C percentage of the infarct area (as a % of the (DCX) or for 3 hours at room temperature (BrdU ipsilateral hemisphere area) or as the corrected or NeuN): rabbit anti-doublecortin antibodies infarct volume (mm3, corrected for edema), as (DCX, 1:1000, ab18723, Abcam, UK), rat described previously (Morancho et al., 2012). anti-BrdU antibodies (BrdU, 1:100, ab6326, Cortical atrophy was evaluated using cresyl violet Abcam, UK) and anti-NeuN antibodies staining at 14 days after tMCAo (as shown in conjugated to Alexa Fluor® 488 (NeuN, 1:200, study B in Figure 1). Briefly, the mice were deeply MAB377X, Merck Millipore, Germany). Alexa anesthetized and then euthanized after the brain Fluor 488 goat anti-rabbit IgG (H+L) and Alexa was perfused with ice-cold 4% PFA. The brains Fluor 647 goat anti-rat IgG (H+L) (1:500, A-1108 were harvested and post-fixed in PFA overnight & A-21247, Invitrogen, USA) were the secondary at 4°C, followed by cryoprotection in 30% sucrose antibodies, which were applied, where required, in PBS until the brains sank. Afterward, the brains at room temperature for 1 hour. The samples were embedded in optimal cutting temperature were then mounted in VectashieldTM with DAPI compound (OCT; Tissue-Tek, Fisher Scientific, (H-1200, Vector Laboratories, USA) to stain the USA) and stored at -80°C until used. cell nuclei. Twelve-μm-thick coronal sections including the lateral ventricles and the hippocampus were To determine the number of neuroblasts collected. Two sections, from bregma +0.02 and migrating from the SVZ to the peri-infarct tissue, bregma -1.70, were selected per animal from the the DCX+/DAPI+ -labeled area in the SVZ and WT and MMP-13 KO ischemic mice. The sections corpus callosum (CC) regions of one slide was were stained with 0.1% freshly filtered cresyl quantified. In addition, triple-labeled cells violet containing 0.2% acetate and 0.15% acetic (DCX+/BrdU+/DAPI+), which indicated the new acid for 15 minutes at 37°C. The areas of the proliferating neuroblasts, were also quantified in ischemic cortex and the area of the ipsilateral the SVZ and peri-infarct areas (three selected hemisphere were measured blindly and the regions of interest (ROIs) in the infarct boundary results are presented as cortex percentage vs. area). The images were obtained at 400X ipsilateral hemisphere. 45 magnification using a confocal laser scanning collected from the supernatant after biological microscope (FV1000, Olympus, Japan) centrifugation at 12,000 rpm for 12 minutes at and analyzed using FV10-ASW software. Finally, 4°C. A MILLIPLEX® map mouse the mature (NeuN+/DAPI+) neurons in the angiogenesis/growth factor magnetic bead panel peri-infarct cortical areas adjacent to the (MAGPMAG, EMD Millipore, Germany) was used damaged tissue were imaged at 100X to detect angiogenesis-related proteins at day 7 magnification and counted using ImageJ software after ischemia. Briefly, 25 μl of the diluted lysates (NIH, USA). The results are presented as the (1:2 in assay buffer) was loaded into the sample neurons per area. All analyses were conducted by wells to quantify the levels of angiopoietin-2, a researcher who was blinded to the genotype granulocyte colony stimulating factor (G-CSF), and group. To study angiogenesis, the mice from amphiregulin, fibroblast growth factor (FGF)-2, study B (see Figure 1) were injected vascular endothelial growth factor-A (VEGF-A) intravenously (retro-orbitally) with Dylight and stromal cell-derived factor- 1α (SDF-1), 594-labeled tomato lectin (80 μg/mouse, according to the manufacturer’s instructions. The DL-1177, Vector Laboratories, USA) 10 minutes assay sensitivity for the different proteins was as before euthanized. Two similar isotopically follows: angiopoietin-2, 5.2 pg/ml; G-CSF, 0.8 oriented slides per animal were selected from pg/ml; amphiregulin, 2.2 pg/ml; FGF-2, 10.6 the brain sections. One image from each section pg/ml; VEGF-A, 0.7 pg/ml and SDF-1, 24.6 pg/ml. that contained the infarct boundary area was Each sample was assayed in duplicate, and the randomly imaged at 100X magnification. coefficients of variation were <20%. The results Angiogenesis was evaluated as the lectin-positive are presented as the ratio of the amount of blood vessel density in the selected ROIs and the protein in the ipsilateral cortex to the amount of corresponding contralateral areas. The total area protein in the contralateral cortex. To evaluate of the lectin-positive perfused vessels was the potential MMP protein imbalance in a calculated using ImageJ software (NIH, USA) by MMP-13 knockdown, a MILLIPLEX® map mouse an investigator blinded to the genotype group. MMP magnetic bead panel 3 (MMMP-3 MAG, EMD Millipore, Germany) was run to detect the Multiplex ELISA expression of MMP-3, MMP-8 and MMP-9 in Both the WT and the MMP-13 KO mice (see brain homogenates obtained at 3 days after study protocol C in Figure 1) were transcardially reperfusion (tissues were homogenized right perfused with ice-cold saline and euthanized at 7 after TTC staining), as described above. In brief, days after ischemia. The mouse brain tissue 25 μl of the brain homogenate (1:10 in lysis obtained from both the ipsilateral and buffer) was loaded into the sample wells. The contralateral cortex was homogenized with assay sensitivity of the different MMPs was freshly prepared ice-cold lysis buffer containing MMP-3, 2.5 pg/ml; MMP-8, 3.4 pg/ml and 50 mM Tris-HCl, 150 mM NaCl, 5 mM CaCl2, 0.05% MMP-9, 3.1 pg/ml. Each sample was assayed in BRIJ-35, 0.02% NaN3, 1% Triton X-100, 1% duplicate, and the coefficients of variation were phenylmethanesulfonyl fluoride (PMSF; <20%. Sigma-Aldrich, Switzerland) and 0.5% aprotinin EPC cultures and MMP-13 silencing (Sigma-Aldrich, USA). Cortical lysates were 46 Outgrowth endothelial cells (OECs) with an (RNase-free water) for the reverse transcription endothelial-like phenotype and a highly to cDNA to detect possible contamination. proliferative nature were obtained from FVB RT-PCR was performed by mixing 5 μl of Taq Man wild-type male mice spleens for a previous study 279 Universal 2X PCR Master Mix (Applied (Morancho et al., 2013) and stored in liquid Biosystems, USA), 0.5 μl of the Taq Man® Gene nitrogen until use. These cells behave as Expression Assay solution (PPIB: Mm competent endothelial cells both in vitro and in 04208118_gl, GAPDH: Mm99999915_g1, vivo (Fadini et al., 2012). MMP-13 gene silencing MMP-13: Mm00439491_m1; Applied Biosystems, was achieved using delivery of passive siRNA via USA), 3.5 μl of RNase-free water and 1 μl of the Accell SMART pool siRNA (Thermo Scientific, USA) sample (cDNA or RNase-free water). These according to the manufacturer’s protocol. Briefly, experiments were performed in triplicate. A 2×104 EPCs were seeded in 48-well plates and sample calibrator was used to compare the treated separately after 24 h with the respective samples from different reading plates. siRNA solutions corresponding to each siRNA Hybridizations and expression analyses were pool (2 μmol/L). Non-targeting/non-coding siRNA performed using a 7900 thermocycler HT Fast was used as a negative control to measure the Real-Time PCR System (Applied Biosystems, USA), minimal toxicity and side effects of the assay. and the results were analyzed using SDS 2.3 Cyclophilin B (PPIB) siRNA was used as a positive software. control to test the efficiency and specificity of MMP-13 silencing. After 48 hours of treatment, In vitro tubulogenesis the cells were thoroughly washed with PBS, To assess the angiogenic and/or vasculogenic trypsinized and collected in basal EBM media ability of MMP-13 silenced cells, MatrigelTM (Lonza, Spain) to perform Matrigel assays, which matrix (BD Biosciences, USA) was used for in vitro are described below. To confirm the mRNA tube formation assays (also named expression in each condition, real-time PCR was tubulogenesis). The experimental groups conducted using the mRNA that was obtained consisted of MMP-13-silenced OECs, from Matrigel-disaggregated cells, which were non-targeting/non-coding siRNA OECs and isolated using the Cell Recovery Solution (BD Cyclophilin B (PPIB) siRNA OECs. Briefly, 9,000 Biosciences, USA) for 60 minutes at 4°C. Briefly, cells were seeded in 96-well plates that were an RNeasy® Minikit (Qiagen, Germany) was used, previously coated with 50 μl of MatrigelTM matrix according to the manufacturer’s instructions, and and then maintained in a CO2 incubator at 37°C. the quantity and quality of the RNA was After 24 hours, two images per well were measured using a Nanodrop Spectrophotometer acquired using an Olympus IX71 microscope and retrotranscribed as cDNA using a High- (100X magnification). Each experiment was run in Capacity cDNA Reverse Transcription Kit (Applied duplicate in four independent experiments. In Biosystems, USA). The incubation conditions for addition, to monitor the formation of vessel-like the retrotranscription were: 10 min at 25°C, 120 structures by mouse EPCs, a time lapse imaging min at 37°C, and 5 min at 85°C. Finally, the assay was performed over hours, as described in samples were incubated at 4°C until use or frozen a previous study (Morancho et al., 2013). A at -80°C (Thermal Cycler 2720, Applied standard MatrigelTM assay was conducted as Biosystems, USA). We included a negative control 47 described above; however, continuous image RESULTS acquisition was performed starting 2 hours after MMP-13 deficiency protects the brain from seeding and then every 30 minutes (for up to 24 ischemia-reperfusion injury and improves hours). An Olympus multi-dimensional-TIRFM functional outcomes cell-R microscope (Olympus, Japan) with temperature, CO2 and humidity control was used. The WT and KO alleles for MMP-13 were Two images per well were acquired at 100X confirmed using PCR (Figure 2A). The brain magnification. For both assays, the number of infarct was localized to the somatosensory and complete rings, the number of branching points motor cortex in both the permanent and the and the total tube lengths (the perimeter of transient model (Morancho et al., 2012). complete rings) were automatically counted Unexpectedly, 4 WT mice presented some infarct blindly using Wimasis® Image Analysis software. expansion that affected the striatum at 3 days The results are expressed as a percentage of the after MCAo. MMP-13 deficiency reduced infarct control condition. volume in the ischemia-reperfusion model at both 24 hours and 3 days (26.1±9.4 mm3 for WT Statistical Analysis vs. 14.4±8.3 mm3 for MMP-13 KO, p=0.025; and 24.8±4.7 mm3 for WT vs. 9.3±3.1 mm3 for SPSS version 15.0 was used for statistical analysis. MMP-13 KO, p<0.001; respectively), as shown in Normality was assessed for continuous variables Figure 2B-C. However, this difference was not using the Shapiro-Wilk test. The values from the significant in the permanent occlusion model normally distributed variables are expressed as (19.1±8.7 mm3 for WT vs. 14.5±6.5 mm3 for the mean ± SD and represented as bar graphs. MMP-13 KO, p=0.228). The difference between The values from the non-normally distributed these two ischemia models was attributed to the variables are expressed as the median increase in infarct extension observed in the WT (interquartile range) and represented as box mice subjected to tMCAo, although this plots. The differences between the different difference did not reach statistical significance at groups were analyzed using an independent 24 hours (19.1±8.7 mm3 for pMCAo vs. 26.1±9.4 t-test and ANOVA and, if significant, followed by mm3 for tMCAo, p=0.13) or at 3 days (19.1±8.7 the Dunnett post hoc test (all vs. the control mm3 for pMCAo vs. 24.8±4.7 mm3 for tMCAo, condition) if normally distributed or the p=0.105). An analysis of functional outcomes also Mann-Whitney U-test and Kruskal-Wallis if not confirmed a significant decrease in forelimb force normally distributed. To assess differences at 24 hours in the WT mice (decreased 83.8±15.8% between time-points, we used repeated ANOVA from baseline, p<0.01); however, this was not followed by Bonferroni post hoc tests. Finally clearly observed in the MMP-13 KO mice despite differences between categorical variables were a statistical trend (90.2±23.9% from baseline, analyzed using Fisher’s exact tests or Chi Square p=0.083); see Figure 2D. No differences were tests. A p-value less than 0.05 was considered observed between mice with different genotypes statistically significant. at either 24 hours or 3 days (data not shown). 48 Figure 2. MMP-13 deficiency protects brain tissue during cerebral ischemia reperfusion. Genotype characterization of wild-type (WT) and MMP-13 knockout (KO) mice: a 1300Kb band indicates a WT allele, and a 1485Kb band indicates a MMP-13 KO allele (A). TTC staining showing infarct areas at 24 h and 3 days after 60 min of transient ischemia (B). Bar graph showing mean ± SD infarct volume at 24 h and 3 days; n=6-8/group, * P<0.05 and ** P<0.01 (C). Grip strength measurements confirming that the significant neurological impairment observed in WT mice was reduced in MMP-13-deficient animals; n=13-14/group, ** P<0.01 vs. pre-tMCAo (D). Interestingly, we observed more spontaneous p=0.015, respectively). When the neurological intracranial hemorrhagic transformations at 3 outcomes within the groups were analyzed, only days than at 24 hours, and this was more the WT mice exhibited a significant decrease in common in the WT (6/6 and 4/8, respectively) forelimb force values vs. baseline at 1 week than in the MMP-13 KO mice (3/6 and 1/7, (66±4.6%, p<0.001) and 2 weeks (70.2±5.7%, respectively), as shown in representative images p=0.017), as shown in Figure 3A. At two weeks, of the WT in Figure 2B. The differences between differences in forelimb force were no longer genotypes or reperfusion day were not observed. statistically significant, however if the only presence of HT was considered, regardless the MMP-13 and cortical damage duration of MCAo, MMP-13 deficiency protected Cerebral ischemia leads to brain damage at the brains from spontaneous hemorrhagic same time that it triggers neurorepair conversions (10/14 in WT vs. 4/13 in MMP-13-KO, mechanisms. Cortical atrophy was evaluated at p=0.035, supporting Table 1 and Figure 1). The 14 days after tMCAo as a percentage of the animals studied over a longer time (two weeks, remaining cortex in the ipsilateral hemisphere study B) showed significant differences in (35.7±1.9% for WT vs. 36.1±1.7% for MMP-13 KO, forelimb force between WT and MMP-13 KO at p=0.684) as shown in Figure 3B, regardless of 24 hours and 1 week after tMCAo (p=0.035 and large differences in baseline infarct volumes. 49 Figure 3. Recovery from brain ischemia reperfusion. Grip strength performance: MMP-13 KO mice were acutely protected from functional impairment following cerebral ischemia. After two weeks, differences # between genotypes were no longer observed. n=8, *P < 0.05, **P < 0.01 vs. pre-tMCAo in the same cohort; P < 0.05 vs. WT at the same time point (A). Cortical atrophy of the ipsilateral hemisphere at 2 weeks, n=8 (B). A reduced number of newly born neuroblasts replicating cells in the post-stroke brain (day 14), was observed in the peri-infarct areas of the we found a reduced number of newly born MMP-13 deficient mice during recovery neuroblasts (DCX+/BrdU+) in the peri-infarct DCX+ neuroblasts/immature neurons and areas of the MMP-13 deficient mice (18.0±2.8 proliferating cells were studied in the post stroke cells in WT vs. 4.7±1.3 cells in MMP-13 KO, brain during recovery (at days 7 and 14), with a p=0.001), although a similar number of DCX+ focus on the SVZ, the corpus callosum (CC) and cells was found in the same area (136±27 cells in peri-infarct areas. At the studied time-points, no WT and 130±20 cells in MMP-13 KO, p=0.873; enhanced neuroblast migration was detected at see Figure 4C-D). An additional analysis of these day 7 compared with the contralateral side in cell counts also showed that the percentage of either the WT or the MMP-13 KO mice (Figure newly born neuroblasts (DCX+/BrdU+) out of the 4A), whereas at day 14, the number of total number of neuroblasts was reduced in the neuroblasts was higher in the ipsilateral CC area peri-infarct areas of the MMP-13 deficient mice than in the contralateral CC in both the WT and (16.9% in WT vs. 4.8% in MMP-13 KO, p=0.005). MMP- 13 KO mice (p=0.011 and p=0.011, Immunostaining for mature neurons (NeuN+) did respectively; see Figure 4B). Interestingly, when not reveal any difference between mice with we analyzed the distribution of the BrdU+ different genotypes in numbers or size (data not shown). 50 Figure 4. Neuroblast migration from the SVZ through the CC to the peri-infarct cortex. DCX+ (neuroblast) area quantification in the dorsolateral ventricle (DL) area and the corpus callosum (CC) area, as indicated, at day 7 (A, n=5-6) or day 14 after tMCAo (B, n=8); **P <0.01. The micrographs confirm peri-infarct neurogenesis (the areas studied in selected ROIs) as co-staining between DCX (green), BrdU (yellow), lectin (red) and DAPI (blue, nuclei) on day 14 after tMCAo (C), the white arrows indicate a newborn neuroblast. The bar graphs show that MMP-13 deficiency decreased the number of DCX+/BrdU+-stained proliferating neuroblasts in the peri-infarct cortex; n=8, **P<0.01 (D). MMP-13 is a key protease during post-stroke we sought to study the functional brain vascular remodeling microvasculature using in vivo endothelial labeling with lectin, as described. Our data show Accelerated angiogenesis in the peri-infarct area that 7 days after cerebral ischemia, no difference is closely related to improved post stroke tissue in peri-infarct vessel density (the percentage of repair (Hayashi et al., 2006). In the present study, 51 the lectin-stained area of the ROIs, as shown in area than the MMP-13 KO mice in the peri-infarct Figure 5A) occurred as a consequence of cortex (3.03±0.41 vs. 2.53±0.51, p=0.049), and ischemic insult in either the WT (3.11±0.35 in the this was also observed, although less extensively, ipsilateral vs. 3.15±0.47 in the contralateral side, in the contralateral cortex, where it did not reach p=0.861) or the MMP-13 KO mice (3.16±0.64 in statistical significance (2.45±0.19% in WT vs. the ipsilateral vs. 3.37±0.60 in the contralateral 2.24±0.20% in MMP-13 KO; p=0.05; Figure 5B-C). side, p=0.578). Nevertheless, on day 14, the WT No differences were observed in the peri-infarct mice presented higher vessel density in the area in the striatum, as expected, in the distal peri-infarct cortex than in the corresponding cortical experimental model of stroke (data not contralateral area (3.03±0.41 vs. 2.45±0.19, shown). Interestingly, we studied whether the p=0.003; Figure 5B-C), whereas the MMP-13 KO number of proliferating neuroblasts (DCX+/BrdU+ mice did not display ischemia-enhanced cells) was related to vessel density and observed angiogenesis in the peri-infarct area (2.53±0.51 a noticeable association between these studied vs. 2.24±0.20, p=0.160; Figure 5B-C). Accordingly, parameters, as shown in the correlation plot in the WT mice presented a more extensive vessel Figure 5D (p=0.056). Figure 5. Brain vessel density after ischemia reperfusion. Representation of the studied areas (indicated by squares), (A). Micrographs of Lectin+ (red) vessels in representative brains; scale bar represents 200 μm (B). Bar graphs representing vessel density in the peri-infarct cortex and the corresponding contralateral area on days 7 and 14 after tMCAo; n=6-8, **P<0.01 and *P<0.05 (C). Correlation plot between vessel density and the number of proliferating neuroblasts in peri-infarct areas on day 14 (D). In agreement with this temporal profile, the obtained from immunodetection assays showed levels of several angiogenesis-related growth that G-CSF (shown as a ratio of factors were examined in the ischemic brain at ipsilateral/contralateral) was higher after day 7, which is just before the increase in vessel ischemia in the WT mice than in the sham group density observed in the peri-infarct tissue. Data (5.1 (1.9-10.5) vs. 1.1 (0.8-1.8), respectively; 52 p=0.003). At the same time-point, the MMP-13 knockdown, the levels of three MMPs MMP-13-deficient ischemic mice presented a (including another collagenase, MMP-8) were lower level of cortical G-CSF than the quantified in the ischemic cortex and the corresponding WT mice (2.5±1.5 vs. 5.2±2.7, corresponding contralateral tissue. Interestingly, p=0.031; see Figure 6A). In agreement with these our results show that the levels of MMP-3, results, the expression of other growth factors MMP-8 and MMP-9 were significantly elevated in associated with vascular remodeling was also the damaged cortex in mice with both genotypes significantly lower in the infarct cortex of compared to the contralateral cortex, as shown MMP-13 KO mice than in the WT mice, for in Figure 6B. Moreover, the amount of these example, VEGF-A (0.9±0.3 vs.1.4±0.3, p=0.007) MMPs was considerable higher in the ipsilateral and amphiregulin (0.9±0.2 vs. 1.1±0.2, p=0.049; cortex of WT mice than in the MMP-13 KO mice: see Figure 6A). Other growth factors, such as 1.21 (0.55-1.78) ng/mg vs. 0.14 (0.09-0.20) angiopoietin-2, FGF-2 and SDF-1 were also ng/mg for MMP-3, p=0.002; 0.8±0.2 ng/mg vs. down-regulated in the MMP-13 KO brains but 0.4±0.1 ng/mg for MMP-8, p=0.001; and 2.4±0.9 these differences were not significant when ng/mg vs. 0.7±0.3 ng/mg for MMP-9, p=0.002. compared to the WT (not shown). To explore These results support the idea that MMP-13 effects on other members of the MMP family of plays a role as a central protease that regulates proteins that might potential compensate for other MMPs after ischemia. Figure 6. Pro-angiogenic factors and MMPs in the cortex after ischemia. Bar graphs representing the ratio of MMP proteins between the ipsilateral (IP) and contralateral (CL) hemispheres in WT and MMP-13 KO ischemic mice at 7 days after ischemia; n=7-8, *P<0.05, **P<0.01 (A). Box plots and bar graphs demonstrating the overexpression of other MMPs in the ipsilateral hemisphere of WT brains and their dysregulation in ## MMP-13-deficient mice at day 3. **p<0.01 and p<0.01 vs. ipsilateral (B). Data are expressed as ng of MMP per mg of total protein. Transient silencing of the MMP-13 gene impairs angiogenic/vasculogenic functions of OECs are tubulogenesis in vitro controlled by MMP-13, an in vitro MatrigelTM assay was conducted after MMP-13 expression To determine whether the was silenced using siRNA. As shown in Figure 7A, 53 introducing MMP-13 siRNA into OEC cell cultures targeting of other genes, such as PPIB, reduced the expression of MMP-13 mRNA by successfully reduced their expression (up to 68.62% compared with the control cells 93.63, p<0.001) without modifying MMP-13 (p=0.001). This effect occurred without expression, as shown in Figure 7A. Nonspecific modifying the expression of housekeeping genes, (non-targeting) siRNA did not modify any of the such as PPIB. Additional experiments were expression patterns of MMP-13 or PPIB. conducted to demonstrate that the precise TM Figure 7. Tubulogenesis network formation in Matrigel assays of mouse OECs after silencing MMP-13 expression. Bar graphs represent the level of MMP-13 and PPIB RNA expressed in each culture condition; n=4, **P<0.01 vs. Control (A). Representative micrographs showing vessel-like structures shaped by OECs in different treatment groups (B). The bar graphs show the quantification of the number of branches, the total tube lengths and the loops of vessel-like structures that were induced in non-silenced, MMP-13-silenced and PPIB-silenced OECs; n=4, *P<0.05 vs. non-targeting siRNA EPCs (C). MatrigelTM functional assays showed that OECs structures, as assessed by the analysis of all that did not endogenously express MMP-13 parameters. The difference in the percentage of formed aberrant tubulogenic networks compared loops (circular structures) reached statistical with all the other tested conditions (see Figure significance (p=0.019), whereas the percentage 7B and supporting Videos 1-4). The total number of branches and the percentage of tube lengths of branching points and loops and the total tube are presented as statistical trends after the lengths of the vessel-like structures were statistical analyses among the groups (p=0.132 quantified. Compared with the non-targeting and p=0.052, respectively), as shown in Figure siRNA (control group), the MMP-13-silenced 7C. OECs showed decreased numbers of vessel-like 54 extracellular matrix, including the basal lamina surrounding brain microvessels, which are DISCUSSION modified during angio-vasculogenesis. Some The present study demonstrates for the first time MMPs have been found to contribute to brain that collagenase-3 (MMP-13) participates in injury (Asahi et al., 2001; Gidday et al., 2005) and tissue damage during cerebral ischemia and is also to orchestrate vascular remodeling simultaneously required for neurovascular (Morancho et al., 2013; Zhao et al., 2006). remodeling during the recovery phases. MMP-13 Knowledge regarding the role of MMP-13 is deficiency reduced infarct size and acutely limited in the context of stroke. diminished the deterioration of functional In our study, both MMP-13 KO and WT mice were outcomes observed after experimental transient subjected to permanent or transient ischemia; cerebral ischemia. However, the absence of however, the infarct size was only significantly MMP-13 also reduced the amount of newly born reduced in the MMP-13- deficient mice in the neuroblasts in the peri-infarct areas during transient model at both 24 hours and 3 days after neurological recovery and altered stroke-induced MCAo, indicating that the role of MMP-13 in angiogenesis by reducing peri-infarct tissue damage is most likely more related to the angiogenesis and suppressing the secretion of reperfusion phase. The infarct volume did not G-CSF, angiopoietin-2 and VEGF-A. Finally, change in the MMP-13-KO mice; however, more endothelial progenitor cell function damage was observed in the WT mice in the (demonstrated in their high potential for vascular transient model than in the permanent model. In remodeling) was abolished when MMP-13 gene addition, we unexpectedly observed expression was silenced in vitro. spontaneous hemorrhagic transformations in our Stroke leads to cell and tissue death within model (especially at day 3), but the knockdown minutes/hours because the brain spontaneously of MMP-13 partially protected mice from activates both injury and repair mechanisms that suffering these complications. Previous results compete for the recruitment of tissue after an from our laboratory revealed a relationship ischemic insult (Lo EH, 2008). This state opens a between a high plasma level of MMP-13 and new battlefield wherein neuroprotection is early brain lesion growth in stroke patients needed to minimize the expansion of the injury (Rosell et al., 2005). We identified an increase in and neurorepair strategies are needed to MMP-13 protein levels in the infarct/peri-infarct enhance endogenous repair. Among the known areas in human and rat tissues, and the neurorepair mechanisms is the process by which activation of neuronal nuclei has been related to cerebral ischemia triggers spontaneous cell death in oxygen- and glucose deprivation neurogenesis (Ohab et al., 2006; Kreuzberg et al., studies (Cuadrado et al., 2009a and 2009b). 2010), angiogenesis (Ergul et al., 2012) and Recently, increased MMP-13 mRNA and protein gliogenesis (Tanaka et al., 2003; Miyamoto et al., levels were observed in brain tissue during the 2010) in peri-infarct areas. From a mechanistic acute phase of ischemia reperfusion by Lenglet point of view, our efforts have been focused on and colleagues (Lenglet et al., 2014). Other MMPs, which are zinc-dependent authors have reported that injured endothelial endopeptidases that degrade components of the cells release MMPs, including MMP-13, which 55 attack the basal lamina and degrade the matrix observed between the genotypes during the first components and tight junctions of endothelial week after ischemia (the WT mice performed cells (Hattori et al., 2003). Ueno and colleagues worse); however, this difference disappeared at 2 found that the expression of MMP-13 in weeks, at which time the WT mice showed hippocampal vessels was increased in rats with improved results in the forelimb force test. These blood brain barrier (BBB)- damaged vessels data support the idea that MMP-13 is implicated compared with rats without BBB impairment, in the post-stroke spontaneous-recovery phase at indicating the importance of MMP-13 a functional level. In this regard, our results also recruitment to brain tissue during the healing showed that similar cortical atrophy occurred in process, where it participates in angiogenesis MMP-13 KO and WT mice at 2 weeks, despite the and neurogenesis (Ueno et al., 2009). Indeed, it larger infarct volumes observed in WT mice, has been demonstrated that a high level of which could be a consequence of different MMP-13 activity lasts for up to 14 days after capabilities related to neurorepair, the inhibition ischemia, and this appears to be closely of cell apoptosis or the promotion of neuronal associated with an increase in aggrecan, survival, among other mechanisms. It will be suggesting their role in neuronal reorganization interesting to perform studies using longer time (Nagel et al., 2005). All these data support a role points to follow-up on these observations for MMP-13 in the acute and subacute phases of regarding the state of atrophy in the damaged stroke: MMP-13 participates in the development tissue. of acute ischemic damage, while at the same time contributing to favorable neurovascular Neurogenesis and angiogenesis were studied to remodeling during recovery. As we will discuss in further explore the mechanisms affected by the detail, the results in this report support these lack of MMP-13 that could influence cerebral hypotheses. plasticity. Our results confirm that at 2 weeks, neuroblasts had migrated to the peri-infarct The distal MCAo model was chosen for this study areas, as has been described to occur in due to its association with low mortality rates post-stroke brains (Arvidsson et al., 2002; and for the precise infarct location it induces in Carmichael, 2008). This effect occurred from the cortical areas, which leads to a defined cortical SVZ through the corpus callosum, as expected, peri-infarct area. However, cortical damage tends and no difference was observed between the to produce milder neurological deficits over the genotypes. Interestingly, the peri-infarct areas of long-term. In our experience, the grip strength the mice lacking MMP-13 contained a reduced test has been proven to be a feasible method for number of proliferating neuroblasts short- and long-term evaluations after distal (DCX+/BrdU+), which are responsible for the MCAo (Rosell et al., 2013b). Our data show that formation of new neural cells in the areas the WT mice presented increased neurological showing active neurorepair. Most neuroblasts die deterioration in the first week. No difference was during migration and the generation of new detected in the MMP-13 KO mice, supporting a neurons (Zhang et al., 2004). We hypothesized protective role for MMP-13 in acutely induced that either the microenvironment in peri-infarct neurological deficits. Moreover, significantly areas and along the migration pathways in brains different effects on neurological functions were lacking MMP-13 did not support neuroblast 56 survival or that lack of MMP-13 may result in brain resident cells, such as astrocytes or mature decelerated standard migration patterns. In this endothelial cells, as has been described by others regard, neurogenesis has been shown to be (Ohab and Carmichael, 2008; Kahle and Bix, coupled to angiogenesis in a post-stroke 2013). However, they could also be released by neurovascular niche (Ohab et al., 2006), and it stem/progenitor cells, such as EPCs or NPCs, has been demonstrated that blood vessels in the which secrete multiple growth factors (Rosell et peri-infarct cortex secrete soluble factors that al., 2013a; Drago et al., 2013; Urbich et al., 2011). maintain neurogenic potential and guide neurogenesis toward the damaged cerebral EPCs have been demonstrated to have cortex (Shen et al., 2004; Thored et al., 2007). In therapeutic potential in that they 626 directly support of these hypotheses, we observed a enhance angiogenesis and neurogenesis positive association between the number of (Asahara et al., 1997; Carmeliet, 2003; Ma et al., proliferating neuroblasts and lectin-positive 2015). These cells, which are also the source of vessels in tissues with more extensive vessel multiple trophic factors, are able to differentiate networks and the ones that presented more into cells with an endothelial phenotype and newborn neuroblasts in the peri-infarct area. have been proven to reduce the brain damage caused by ischemia and to enhance Importantly, our study shows that at day 7 after angio-vasculogenesis in stroke models (Ohta et ischemia-reperfusion injury, lower amounts of al., 2006; Moubarik et al., 2011; Rosell et al., some nourishing factors, such as G-CSF, 2013a). In the present study, we further angiopoietin-2 and VEGF-A, were observed in the demonstrated a role for MMP-13 in tubulogenic infarct cortex of the MMP-13 KO mice than in the remodeling in experiments in which we silenced WT brains. This is an interesting finding because the expression of MMP-13 in EPCs, which have the administration of some factors in the early been shown to be involved in angiognesis stages of stroke has been shown to stimulate (Morancho et al., 2013; Rosell et al., 2013a). Our angio-neurogenesis and to lead to improved data demonstrate that EPCs develop abnormal functional recovery (Schäbitz et al., 2003; Rivera tubulogenic remodeling patterns when MMP-13 and Bergers, 2014; Beck et al., 2000). These data expression is blocked, in which they failed to suggest that MMP-13 could be indirectly shape vessel-like networks in MatrigelTM contributing to an appropriate angiogenic substrate. This interesting observation should be micro-environment. Therefore, we expected to considered for future therapeutic strategies, observe changes in vessel density between the although additional research in an in vivo setting MMP-13 and WT mice. In agreement with this is needed to study the effects of MMP-13 assumption, our observations support a role for deficiency in tissue EPCs or in putative EPC-based MMP-13 in vascular remodeling after stroke: two cell therapies. weeks after cerebral ischemia, increased vessel density in the peri-infarct area was found only in In summary, our study highlights the role of the WT animals and not in the MMP-13-deficient MMP-13, a pivotal collagenase that remodels the mice. The present study does not identify the ECM, in neuroprotection and neurorepair after source of the trophic factors that are affected by cerebral ischemia. This central role is further MMP-13 deficiency. These could be secreted by supported by our findings regarding the 57 regulation of other MMPs in the ischemic brain. Therefore, lacking MMP-13 protects the brain We demonstrate that in WT mice, other MMPs, and improves functional outcomes over the short including MMP-3, MMP-8 and MMP-9, are term, whereas over longer periods of time, this upregulated in the ischemic brain, as was protease is required to support vascular reported by our group in human infarct tissue remodeling and the production of growth factors (Cuadrado et al., 2009b), but also that MMP-13 that are linked to peri-infarct neurogenesis. In deficiency considerably diminished the conclusion, the therapeutic modulation of overexpression of these other MMP family MMP-13 during stroke is fully supported by the members. present data. ACKNOWLEDGMENTS FM holds a fellowship from the China Scholarship network INVICTUS (RD12/0014/0005). The Council (CSC), and AR is supported by the Miguel authors declare no competing financial interests. Servet program (CP09/00265) from the Spanish We are grateful to Dr. López-Otín from the Ministry of Health (Instituto de Salud Carlos III). Universidad de Oviedo for providing MMP-13 WT This work was supported by research grants from and KO mice. the Fondo de Investigaciones Sanitarias (PI13/01094) and the Spanish stroke research REFERENCES Arvidsson A, Collin T, Kirik D, Kokaia Z, Lindvall O angiopoietin-2, and tie receptors after middle (2002) Neuronal replacement from endogenous cerebral artery occlusion in the rat. Am J Pathol precursors in the adult brain after stroke. Nat 157(5):1473-1483. Med 8: 963-970. Bovetti S, Bovolin P, Perroteau I, Puche AC (2007) Asahara T, Murohara T, Sullivan A, Silver M, van Subventricular zone-derived neuroblast der Zee R, Li T, Witzenbichler B, Schatteman G, migration to the olfactory bulb is modulated by Isner JM (1997) Isolation of putative progenitor matrix remodelling. Eur J Neurosci 25: endothelial cells for angiogenesis. Science 2021-2033. 275(5302): 964-967. Carmeliet P (2003) Angiogenesis in health and Asahi M, Wang X, Mori T, Sumii T, Jung JC, disease. Nat Med 9: 653-660. Moskowitz MA, Fini ME, Lo EH (2001) Effects of Carmicheal ST (2008) Themes and strategies for matrix metalloproteinase-9 gene knock-out on studying the biology of stroke recovery in the the proteolysis of blood-brain barrier and white post-stroke epoch. Stroke 39(4): 1380-1388. matter components after cerebral ischemia. J Neurosci 21(19): 7724-7732. Cuadrado E, Rosell A, Borrell-Pagès M, García-Bonilla L, Hernández-Guillamon M, Beck H, Acker T, Wiessner C, Allegrini PR, Plate Ortega-Aznar A, Montaner J (2009a) Matrix KH (2000) Expression of angiopoietin-1, metalloproteinase-13 is activated and is found in 58 the nucleus of neural cells after cerebral ischemia. wound healing. Am J Pathol 175(2): 533-546. J. Cereb. Blood Flow Metab 29: 398-410. Hayashi T, Deguchi K, Nagotani S, Zhang H, Cuadrado E, Rosell A, Penalba A, Slevin M, Sehara Y, Tsuchiya A, Abe K (2006) Cerebral Alvarez-Sabin J, Ortega-Aznar A, Montaner J ischemia and angiogenesis. Curr Neurovasc Res (2009b) Vascular MMP-9/TIMP-2 and neuronal 3(2): 119-129. MMP-10 up-regulation in human brain after Hayakawa K, Miyamoto N, Seo JH, Pham LD, Kim stroke: a combined laser microdissection and KW, Lo EH, Arai K (2013) High mobility group box protein array study. J Proteome Res 8(6): 1 from reactive astrocytes enhances the 3191-3197. accumulation of endothelial progenitor cells in Drago D, Cossetti C, Iraci N, Gaude E, Musco G, damaged white matter. Neuroscience 125(2): Bachi A, Pluchino S (2013) The stem cell 273-280. secretome and its role in brain repair. Biochimie Inada M, Wang Y, Byrne MH, Rahman MU, 95(12): 2271-2285. Miyaura C, López-717 Otín C, Krane SM (2004) Ergul A, Alhusban A, Fagan SC (2012) Critical roles for collagenase-3 (Mmp13) in Angiogenesis: A harmonized target for recovery development of growth plate cartilage and in after stroke. Stroke 43: 2270-2274. endochondral ossification. Proc Natl Acad Sci USA 101(49): 17192-17197. Fadini GP, Losordo D, Dimmeler S (2012) Critical reevaluation 695 of endothelial progenitor cell Jovin TG et al. REVASCAT Trial Investigators (2015) phenotypes for therapeutic and diagnostic use. Thrombectomy within 8 Hours after Symptom Circ Res 110(4): 624-637. Onset in Ischemic Stroke. N Engl J Med. 2015 372(24):2296-306. Gidday JM, Gasche YG, Copin JC, Shah AR, Perez RS, Shapiro SD, Chan PH, Park TS (2005) Kahle MP, Bix GJ (2013) Neuronal restoration Leukocyte-derived matrix metalloproteinase-9 following ischemic stroke: influences, barriers, mediates blood-brain barrier breakdown and is and therapeutic potential. Neurorehabil Neural proinflammatory after transient focal cerebral Repai 27(5): 469-478. ischemia. Am J Physiol Heart Circ Physiol 289: Knäuper V, Cowell S, Smith B, López-Otin C, H558-H568. O'Shea M, Morris H, Zardi L, Murphy G (1997) Hacke W, Kaste M, Bluhmki E, Brozman M, The role of the C-terminal domain of human Davalos A, Guidetti D, Larrue V, Lees KR, collagenase-3 (MMP-13) in the activation of Medeghri Z, Machnig T, Schneider D, von procollagenase-3, substrate specificity, and tissue Kummer R, Wahlgren N, Toni D; ECASS inhibitor of metalloproteinase interaction. J Biol Investigators (2008) Thrombolysis with alteplase Chem 272(12): 7608-7616. 3 to 4.5 hours after acute ischemic stroke. N Engl Knäuper V, López-Otin C, Smith B, Knight G, J Med 359: 1317-1329. Murphy G (1996) Biochemical characterization of Hattori Y, Nerusu KC, Bhagavathula N, Brennan M, human collagenase-3. J Biol Chem 271(3): Hattori N, Murphy HS, Su LD, Wang TS, Johnson 1544-1550. TM, Varani J (2003) Vascular expression of matrix Kreuzberg M, Kanov E, Timofeev O, Schwaninger metalloproteinase-13 (collagenase-3) in basal cell M, Monyer H, Khodosevich K (2010) Increased carcinoma. Exp Mol Pathol 74(3): 230-237. subventricular zone-derived cortical Hattori N, Mochizuki S, Kishi K, Nakajima T, neurogenesis after ischemic lesion. Exp Neurol Takaishi H, D’Armiento J, Okada Y (2009) MMP-13 226: 90-99. plays a role in keratinocyte migration, Lees KR et al. (2010) Time to treatment with angiogenesis, and contraction in mouse skin intravenous alteplase and outcome in stroke: an 59 updated pooled analysis of ECASS, ATLANTIS, Pisano P (2011) Transplanted late outgrowth NINDS, and EPITHET trials. Lancet 375: endothelial progenitor cells as cell therapy 1695-1703. product for stroke. Stem Cell Rev and Rep 7: 208-220. Lenglet S, Montecucco F, Mach F, Schaller K, Gasche Y, Copin JC (2014) Analysis of the Nagel S, Sandy JD, Meyding-Lamade U, Schwark C, expression of nine secreted matrix Bartsch JW, Wagner S (2005) Focal cerebral metalloproteinases and their endogenous ischemia induced changes in both MMP-13 and inhibitors in the brain of mice subjected to aggrecan around individual neurons. Brain Res ischaemic stroke. Thromb 739 Haemost 112(2): 1056 (1): 43-50. 363-378. Ohab JJ, Fleming S, Blesch A, Carmichael ST (2006) Lo EH (2008) A new penumbra: transitioning A neurovascular niche for neurogenesis after from injury into repair after stroke. Nat Med stroke. J Neurosci 26 (50): 13007-13016. 14(5): 497-500. Ohab JJ, Carmichael ST (2008) Poststroke Ma F, Morancho A, Montaner J, Rosell A (2015) neurogenesis: emerging principles of migration Endothelial progenitor cells and revascularization and localization of immature neurons. following stroke. Brain Res 1623: 150-159. Neuroscientist 14(4): 369-380. Miyamoto N, Tanaka R, Shimura H, Watanabe T, Ohta T, Kikuta K, Imamura H, Takagi Y, Nishimura Mori H, Onodera M, Mochizuki H, Hattori N, M, Arakawa Y, Hashimoto N, Nozaki K (2006) Urabe T (2010) Phosphodiesterase III inhibition Administration of ex vivo-expanded bone promotes differentiation and survival of marrow-derived endothelial progenitor cells oligodendrocyte progenitors and enhances attenuates focal cerebral ischemia-reperfusion regeneration of ischemic white matter lesions in injury in rats. Neurosurgery 59(3): 679-686. the adult mammalian brain. J Cereb Blood Flow Rivera LB, Bergers G (2014) Angiogenesis. Metab 30(2): 299-310. Targeting vascular sprouts. Science 344 (6191): Ming G, Song H (2011) Adult Neurogenesis in the 1449-1450. mammalian brain: significant answers and Rosell A, Alvarez-Sabín J, Arenillas JF, Rovira A, significant questions. Neuron 70(4): 687-702. Delgado P, Fernández-Cadenas I, Penalba A, Morancho A, García-Bonilla L, Barceló V, Giralt D, Molina CA, Montaner J (2005) A matrix Campos-Martorell M, Garcia S, Montaner J, metalloproteinase protein array reveals a strong Rosell A (2012) A new method for focal transient relation between MMP-9 and MMP-13 with cerebral ischemia by distal compression of the diffusion-weighted image lesion increase in middle cerebral artery. Neuropathol App human stroke. Stroke (7): 1415-1420. Neurobio 38: 617- 627. Rosell A, Lo EH (2008) Multiphasic roles for Morancho A, Hernandez-Guillamon M, Boada C, matrix metalloproteinases after stroke. Curr Opin Barceló V, Giralt D, Ortega L, Montaner J, Rosell A Pharmacol 8(1): 82-89. (2013) Cerebral ischaemia and matrix Rosell A, Morancho A, Navarro-Sobrino M, metalloproteinase-9 modulate the angiogenic Martínez-Saez E, Hernández-784 Guillamon M, function of early and late outgrowth endothelial Lope-Piedrafita S, Barceló V, Borrás F, Penalba A, progenitor cells. J Cell Mol Med 17 (12): García-Bonilla L, Montaner J (2013a) Factors 1543-1553. secreted by endothelial progenitor cells enhance Moubarik C, Guillet B, Youssef B, Codaccioni JL, neurorepair responses after cerebral ischemia in Piercecchi MD, Sabatier F, Lionel P, Dou L, mice. PloS One 8(9): e73244. Foucault-Bertaud A, Velly L, Dignat-George F, Rosell A, Agin V, Rahman M, Morancho A, Ali C, 60 Koistinaho J, Wang X, Vivien D, Schwaninger M, hypertensive model. Hypertens Res 32: 332-338. Montaner J (2013b) Distal occlusion of the Urbich C, De Souza AI, Rossig L, Yin X, Xing Q, middle cerebral artery in mice: are we ready to Prokopi M, Drozdov I, Steiner M, Breuss J, Xu Q, assess long-term functional outcome? Transl Dimmeler S, Mayr M (2011) Proteomic Stroke Res 4 (3): 297-307. characterization of human early pro-angiogenic Schäbitz WR, Kollmar R, Schwaninger M, Juettler cells. J Mol Cell Cardiol 50(2): 333-336. E, Bardutzky J, Schölzke MN, Sommer C, Schwab Ueno M, Wu B, Nishiyama A, Huang CL, Hosomi S (2003) Neuroprotective effect of granulocyte N, Kusaka T, Nakagawa 806 T, Onodera M, Kido M, colony-stimulating factor after focal cerebral Sakamoto H (2009) The expression of matrix ischemia. Stroke 34(3): 745-751. metalloproteinase-13 is increased in vessels with Senger DR, Davis GE (2011) Angiogenesis. Cold blood-brain barrier impairment in a stroke-prone Spring Harb Perspect Biol 3(8): a005090. hypertensive model. Hypertens Res 32: 332-338. Shen Q, Goderie SK, Jin L, Karanth N, Sun Y, Urbich C, De Souza AI, Rossig L, Yin X, Xing Q, Abramova N, Vincent P, Pumiglia K, Temple S Prokopi M, Drozdov I, Steiner M, Breuss J, Xu Q, (2004) Endothelial cells stimulate self-renewal Dimmeler S, Mayr M (2011) Proteomic and expand neurogenesis of neural stem cells. characterization of human early pro-angiogenic Science 304(5675): 1338-1340. cells. J Mol Cell Cardiol 50(2): 333-336. Tanaka K, Nogawa S, Suzuki S, Dembo T, Kosakai Zhang ZG, Chopp M (2009) Neurorestorative A (2003) Upregulation of oligodendrocyte therapies for stroke: underlying mechanisms and progenitor cells associated with restoration of translation to the clinic. Lancet Neurol 8 (5): mature oligodendrocytes and myelination in 491-500. peri-infarct area in the rat brain. Brain Res 989(2): Zhang R, Zhang Z, Zhang C, Zhang L, Robin A, 172-179. Wang Y, Lu M, Chopp M (2004) Stroke transiently Thored P, Wood J, Arvidsson A, Cammenga J, increases subventricular zone cell division from Kokaia Z, Lindvall O (2007) Long-term neuroblast asymmetric to symmetric and increases neuronal migration along blood vessels in an area with differentiation in the adult rat. J Neurosci 24(25): transient angiogenesis and increased 5810-5815 vascularization after stroke. Stroke 38(11): Zhao BQ, Wang S, Kim HY, Storrie H, Rosen BR, 3032-3039. Mooney DJ, Wang X, Lo EH (2006) Role of matrix N, Kusaka T, Nakagawa 806 T, Onodera M, Kido M, metalloproteinases in delayed cortical responses Sakamoto H (2009) The expression of matrix after stroke. Nat Med 12 (4): 441-445. metalloproteinase-13 is increased in vessels with blood-brain barrier impairment in a stroke-prone 61 SUPPORTING INFORMATION Supporting Table 1. Supporting Table 1. Summary of HT incidences in MMP-13 KO and WT mice after transient MCAo. HT incidence 24 hours post-ischemic 72 hours post-ischemic P value reperfusion reperfusion WT (10/14) 50% (4/8) 100% (6/6) 0.085 MMP-13 KO (4/13) 14.3% (1/7) 50% (3/6) 0.266 0.035 0.282 0.182 Supporting Figure 1. Supporting Figure 1. Representative images of HT at 24h or 72h after cerebral ischemia in WT or MMP-13 KO mice. 62 TM Supporting video 1. 24 hours of time-lapse imaging Matrigel assay of mouse WT control OECs (100X) http://ees.elsevier.com/ynbdi/download.aspx?id=341324&guid=a0d8c941-d09c-4535-b9dc-4c39268a 43ca&scheme=1 TM Supporting video 2. 24 hours of time-lapse imaging Matrigel assay of mouse WT MMP-13 silenced OECs (100X) http://ees.elsevier.com/ynbdi/download.aspx?id=341325&guid=03bb1fa4-322e-4b11-9d35-6bd4a99d f1e8&scheme=1 TM Supporting video 3. 24 hours of time-lapse imaging Matrigel assay of mouse WT PPIB silenced OECs (100X) http://ees.elsevier.com/ynbdi/download.aspx?id=341326&guid=12a5e1be-4f0d-4de9-9547-f115b74c 6ad7&scheme=1 TM Supporting video 4. 24 hours of time-lapse imaging Matrigel assay of mouse WT No-targeting silenced OECs (100X) http://ees.elsevier.com/ynbdi/download.aspx?id=341327&guid=0d57cc95-fd0d-4708-94de-058d8253 d2da&scheme=1 63 64 Article 3 Plasma matrix metalloproteinase levels in stroke patients during intensive rehabilitation therapy. Manuscript in preparation. 65 66 Plasma matrix metalloproteinase levels in stroke patients during intensive rehabilitation therapy Feifei Ma1, BSc; Susana Rodriguez2, MD; Xavi Buxo2, MD; Anna Morancho1, PhD; Iolanda Riba-Llena1, MD, PhD; Ana Carrera2; Alejandro Bustamante1, MD; Dolors Giralt1, BSc; Joan Montaner1, MD, PhD; Carmen Martinez2, MD; Immaculada Bori2, MD; Anna Rosell1, PhD. 1 Neurovascular Research Laboratory and Neuroscience Department, Vall d’Hebron Research Institute, Universitat Autonoma de Barcelona, Spain. 2Unidad de Rehabilitacion Neurologica y Dano Cerebral, Hospital Vall d’Hebron, Universitat Autonoma de Barcelona, Spain. Corresponding Author: Anna Rosell, anna.rosell@vhir.org Neurovascular Research Laboratory, Vall d’Hebron Research Institute, Passeig Vall d’Hebron 119-129, 08035, Barcelona, Spain. Tel. +34 934894029 ABSTRACT levels remained stable during the study period and were similar to those in controls. However, baseline MMP-12 and MMP-13 levels were Background: Rehabilitation therapies are still the strongly associated with stroke severity as only approved treatments used to improve measured by NIHSS score (p<0.001 and p=0.008) neurological recovery in disabled stroke survivors. and were elevated in the patients with the most The use of biomarkers to monitor and predict extensive infarcts (p=0.009 and p=0.058). patient’s recovery could help to adjust individual Interestingly, plasma MMP-3 was independent of rehabilitation programs and to identify baseline stroke characteristics but was found to mechanisms of repair. Objective: To study plasma be increased in patients with better levels of matrix metalloproteinases (MMPs) as motor/functional recovery and in patients with potential markers of recovery during intensive larger improvements during rehabilitation. rehabilitation therapy (IRT) after stroke. Methods: Conclusions: MMPS might act as biological Patients with first-ever ischemic strokes enrolled markers of recovery during rehabilitation therapy to IRT (≥3 hours per day/5 days per week) were related to their roles in both injury and tissue studied (n=15). The protocol consisted of a remodeling. Future confirmatory investigations in battery of motor/functional tests (Rankin, BI, multicenter studies are warranted by our data. FMA, FAC, MRC, CAHAI and the 10 meter walk) performed before therapy and at one-, three and six-month follow-ups, in addition to blood extractions to measure MMP-3, MMP-12 and Key Words: Stroke, matrix metalloproteinase, MMP-13 levels. Healthy volunteers served as intensive rehabilitation therapy, biomarker, non-ischemic controls (n=15). Results: MMP recovery 67 repair after stroke (8, 9). We have investigated MMP-3, MMP-12 and MMP-13 as they were INTRODUCTION previously related to ischemic brain injury but not to recovery after stroke. Our hypothesis is Despite enormous advances in acute stroke that some of these MMPs can serve as management during the last two decades (1), biomarkers to monitor individual neurological approximately five million people survive a stroke status and/or functional improvement as a each year worldwide with disabilities that limit reflection of the brain plasticity phenomena and independence for several types of daily activities. an individual response to rehabilitation therapy. For these patients, the only approved treatment during the sub-acute and chronic phases of the We designed a prospective study of ischemic disease is neurorehabilitation, with the objective stroke patients who were enrolled in intensive of improving their independency status and to rehabilitation therapy (IRT) to measure plasma reach the best quality of life at long-term (2, 3). levels of MMPs and to determine the motor/functional status of the patients by using During rehabilitation, improved outcomes in a battery of tests up to six months after stroke. patients are the result of early mobilization, rehabilitation intensity and other measures used to prevent and treat medical complications (4, 5), and also to individual responses to therapy. From METHODS a biological perspective, we know that Study Subjects neurorehabilitation enhances different aspects of brain plasticity, such as axonal remodeling, cell The study cohort comprised patients who were genesis and angiogenesis (6), but this could be enrolled in IRT at the Vall d’Hebron University patient-dependent, and the biological Hospital from February 2014 to February 2015. mechanisms responsible for these individual Inclusion criteria were: first-ever ischemic stroke, responses to treatment are still being age < 70 years, somatosensory or ataxic investigated. With this background, it is hemiparesis, time until start of IRT < 3 weeks necessary to investigate the precise molecular after stroke, stable medical condition, endurance mechanisms underlying tissue remodeling and to participate in a minimum of 3 hours/day in a plasticity to ameliorate or adjust current therapy program and agreement to participate in rehabilitation programs. Additionally, the the study. Exclusion criteria included sensory or accessibility of biomarkers that could support an global aphasias, cognitive deficits (Mini-Mental accurate and individual prognosis of recovery State Examination < 23), terminal illness, during rehabilitation would help practitioners to inflammatory diseases or previous deficits of the decide on the intensity, type or duration of the upper/lower limb. In parallel, fifteen program and to personalize treatments (7). non-ischemic healthy individuals with no known neurological, malignant or inflammatory disease In this investigation, we have focused on matrix volunteered to participate as part of the control metalloproteinases (MMPs), enzymes that can cohort. Part of the control cohort were degrade most components of the extracellular hypertensive subjects also studied in the YSSIS matrix that are involved in both injury and tissue 68 study (10) free of ischemic events. criteria for an intensive rehabilitation program were transferred to the Neurorehabilitation Unit The Study Protocol was presented and approved for inpatient IRT, whereas those patients who by the local Clinical Research Ethics Committee were able to walk with some support/assistance (PRIR 317/2013). All patients and controls signed were released to their household and started the the corresponding informed consent. daily IRT at the Day hospital of the Neuro-Rehabilitation Unit. With the posterior Rehabilitation Intervention release from the inpatient Neurorehabilitation Patients who met the inclusion criteria followed a Unit, the IRT then continued at the Day Hospital comprehensive intensive rehabilitation program, until completion of a minimum of 75% of the including physiotherapy, occupational therapy, proposed objectives or when functional stability speech therapy and/or neuropsychology, if was achieved. If there were further objectives, required, for a minimum of 3 hours per day and 5 the patient continued a high-moderate intensity days per week, as indicated by the Catalan outpatient rehabilitation program until the clinical practice guidelines for the management achievement of functional stability. of stroke (11). The protocol started with a physiatrist assessing patient’s deficits within the Study Protocol first 24 hours after stroke and designing an initial A total of fifteen patients were finally included in rehabilitation program. Passive mobilizations the study. Of these, one patient was excluded performed in bed started within the first 24 following the baseline visit after having a second hours after stroke if there are no medical stroke, while another patient voluntarily conditions that contraindicate it, and at 48 hours discontinued his participation. Two other patients started a moderate intensity patients chose to withdraw from the study after rehabilitation program. Those patients who the one-month follow-up visit (see Figure 1). presented severe or moderate deficits in two or more functional areas and who met the inclusion Figure 1. Study flow diagram. 69 Data related to subject demographics, risk factors, “cannot walk” (score 0), “dependent walk” (score medication, comorbidities and exercise were 1-3) and “independent walk” (score 4-5), and obtained from patients and controls (see Table 1), improvement was defined as a shift to an upper along with clinical stroke characteristics (see category (17). For MRC, our analysis Table 2). differentiated between normal (score 5) or impaired (score 0 to 4) muscle strength. Specific visits for the study were performed by an experienced physiatrist at baseline prior to Plasma MMP measure starting the IRT and at one, three and six months Plasma levels of soluble MMP-3, MMP-12 and after starting the therapy program (see Figure 1). MMP-13 were measured with the MILLIPLEXR During these visits, several tests and scales were map Human MMP Panel 1 (HMMP1MAG, EMD performed to assess motor function and Millipore, Germany) for plasma samples obtained functional status (see details below), and blood at baseline (n=15), 1 month (n=13), 3 months was extracted in EDTA tubes and centrifuged at (n=11) and 6 months (n=8). Briefly, 25 µl of 1500 rpm for 15 minutes to obtain plasma, which undiluted plasma were loaded per well and the was stored at −80 °C until use. assay was run following the manufacturer’s Functional/Motor Measures Assessment and instructions. The assay sensitivity for the Definition of improvement different proteins was: MMP-3, 0.062 ng/ml; MMP-12, 0.038 ng/ml and MMP-13, 0.034 ng/ml. The battery of tests consisted of the modified Each sample was assayed in duplicate and only Rankin scale (mRS, 0-6), the Granger's modified values with a CV < 20% were accepted for the BI (12) (BI, 0-100), the Fugl-Meyer Assessment posterior analysis. score for the upper extremity (FMA, 0-66), the Functional Ambulation Categories (FAC, 0-5), the Statistical analysis Chedoke Arm and Hand Activity Inventory (CAHAI, 13-91) (13), the 10-meter walk test and the Descriptive statistics were used to define Medical Research Council (MRC) (scale 0-5) of demographics, stroke characteristics, MMP levels the disabled hemisphere (upper and lower and functional measures. Chi-square tests were extremities at the proximal/distal level). run for categorical variables. Normality of continuous variables was assessed by the Improvement classifications were obtained after Kolmogorov-Smirnov test if N ≥ 30 or the comparing the scores during follow-up visits Shapiro-Wilk test if N < 30. To assess differences versus baseline scores: in FMA was defined as an between independent groups, independent increase ≥ 10 points, described previously as the t-tests and one-way ANOVA followed by minimal clinical important difference (14). For Bonferroni post-hoc tests were used for normally the CAHAI, an improvement was defined as an distributed variables, and the values were increase of ≥ 7 points (15). For the 10-meter expressed as the mean ± s.d., while the Walk Test, the walking velocity was calculated Mann-Whitney U-test was used to explore and improvement was considered if walking differences in non-normally distributed variables velocity increased by more than 0.3 m/sec (16). and was expressed as median and interquartile The FAC was categorized into three categories: range (IQR). For repeated measures, Friedman 70 followed by Wilcoxon tests were used for Briefly, stroke patients were significantly younger non-normally distributed variables, whereas for and had smaller body mass indexes than controls variables with a normal distribution, repeated (p=0.008 and p=0.048, respectively). Males were ANOVA followed by the Bonferroni post hoc test overrepresented in the stroke group (80%) but was run. Correlations between continuous not significantly. All other registered variables variables were measured using the Pearson’s or were balanced between groups. The main clinical Spearman’s coefficients, depending on their variables related to the ischemic event are normality. Finally, a p value <0.05 was considered described in Table 2, which shows common statistically significant. All analyses were features of the cohort of ischemic stroke completed using SPSS 15.0 software. candidates for IRT. In summary, median NIHSS on admission was 8 (5-12), and after 3-4 days it was 5 (5-9), indicating an improvement (increase ≥ 4 points) in 33.3% of the patients. Most strokes RESULTS occurred in the carotid territory (73.3%), and Characteristics of the stroke and control cohorts almost half of these were classified as total anterior cerebral infarction (TACI, (46.7%)). Time Table 1 shows baseline characteristics of the from stroke onset to IRT was 11.4 ± 4.4 days. stroke rehabilitation cohort compared to controls. Table 1. Baseline characteristics of the control and stroke cohorts. Control Strokes P value (n=15) (n=15) Age (years) 63.5±7.6 53.9±10.6 0.008 Gender, male 46.7 (7) 80.0 (12) 0.058 Risk factors Alcohol 53.3 (8) 20.0 (3) 0.058 Tobacco 13.3 (2) 33.3 (5) 0.195 Hypertension 80.0 (12) 66.7 (10) 0.409 Dyslipidemia 60.0 (9) 53.3 (8) 0.713 Diabetes mellitus 26.7 (4) 20.0 (3) 0.666 Atrial fibrillation 0 (0) 13.3 (2) 0.143 Obesity 60.0 (9) 33.3 (5) 0.143 Body mass index (kg/m²) 25.9±3.6 23.0±3.8 0.048 Comorbidities Osteoarticular disorders 26.7 (4) 13.3 (2) 0.361 Ischemic cardiopathy 0 (0) 6.7 (1) 0.309 Psychiatric disorders 26.7 (4) 20.0 (3) 0.666 Previous Exercise 71 Physical activity 73.3 (11) 50.0 (7) 0.196 Physical activity (hours) 7 (0-10) 7 (0-10) 0.288 Previous Medication Antiplatelets 26.7 (4) 20.0 (3) 0.666 Anticoagulants 0 (0) 6.7 (1) 0.309 Statins 33.3 (5) 33.3 (5) 1 Antihypertensives 73.3 (11) 46.7 (7) 0.136 Antidiabetic 20 (3) 13.3 (2) 0.624 Data is displayed as mean ± SD, median (IQR) or percentage % and (number). Significant p values are highlighted in bold. Table 2: Clinical characteristics of stroke patients Stroke Characteristics Strokes (n=15) NIHSS score at admission 8 (5-12) NIHSS motor score at admission 4 (2-7) NIHSS score after 3-4 days 5 (5-9) NIHSS motor score after 3-4 days 3 (1-5) Early neurological outcome Improvement 33.3 (5) Stability 53.4 (8) Worsening 13.3 (2) Etiology Lacunar 13.3 (2) Cardiopathy 33.3 (5) Atherothrombosis 13.3 (2) Others 13.3 (2) Unknown 26.8 (4) Location, Territory Vertebrobasilar 26.7 (4) Carotid 73.3 (11) Hemisphere of stroke Right 53.3 (8) Left 46.7 (7) OCSP classification TACI 46.7 (7) LACI 20.0 (3) 72 PACI 12.3 (2) POCI 20.0 (3) Symptomatic stenosis Extracranial 26.7 (4) Intracranial 20.0 (3) Both 6.7 (1) No 46.6 (7) Asymptomatic stenosis Yes 33.3 (5) No 66.7 (10) Thrombolytic therapy Yes 26.7 (4) No 73.3 (11) Hemorrhagic Transformation Yes 20 (3) No 80 (12) Time Stroke-Intensive Rehabilitation (days) 11.4±4.4 OCSP: Oxfordshire Community Stroke Project; TACI: total anterior cerebral infarct, LACI: lacunar cerebral infarct, PACI: parcial anterior cerebral infarct, POCI: posterior cerebral infarct; Early Neurological Outcome was defined as improvement (increase ≥4 points in the NIHSS), worsening (decrease ≥4 points in the NIHSS) or stability any other change in the NIHSS. Data is shown as mean ± SD, median (IQR) or percentage % (number). Evolution of motor/functional outcomes and and was maintained during the complete MMP levels over time follow-up period. At three months, the patients also showed some improvement in the FAC, The complete motor and functional progress of CAHAI and 10-meter walk test. Patients showed patients during IRT is detailed in Table 3. A minor improvements in MRC scores for the significant improvement in the FMA, Rankin and proximal segment of the extremities at three or BI tests had already been achieved at 1 month six months. Table 3: Measures of functional/motor outcome and MMP profile Strokes Time course Baseline 1 month 3 months 6 months Outcome assessment Fugl Meyer Assessement (0-66) 30 (6-52) 55 (14-61.5)* 60 (16-63)* 64 (26-66)* Modified RANKIN (0-6) 4 (3-4) 2 (1-3)* 1.5 (0-3)** 1 (0-2)* Barthel Index (0-100) 57 (31-83) 94 (64-100)** 100 (88.6-100)** 100 (100-100)** Functional Ambulation Categories (0-5) 1 (0-3) 4 (1.5-5)† 5 (5-5)* 5 (5-5)* 73 Chedoke Arm and Hand Activity Inventory 15 (13-71) 72 (13-85)† 82 (13-90) 88 (13-91) (13-91) 10 meters walk test (velocity in m/sec) 0.5 (0-1) 0.9 (0.3-1.4)† 1 (0.8-1.3) 1.1 (0.8-1.4) Stroke Specific Quality Of Life Scale n.a. 168 (152.5-183) 193 (159-227) 199 (171-229)$ (49-245) MRC Superior-Proximal (0-5) 4 (0-4) 4 (2.5-5) 5 (2-5)* 5 (4-5)* MRC Superior-Distal (0-5) 4 (0-4) 4 (0-5) 5 (0-5) 5 (2-5) MRC Inferior-Proximal (0-5) 4 (0-4) 5 (4-5) 5 (5-5)† 5 (5-5)† MRC Inferior-Distal (0-5) 4 (0-5) 5 (2-5) 5 (4-5) 5 (1-5) MMP level (ng/mL) MMP-3 28.2±14.8 23.6±8.8 32.5±13.4 27.7±12.3 MMP-12 2.0±0.9 2.0±1.0 2.1±0.9 2.1±0.8 MMP-13 4.5±1.7 4.5±2.6 4.3±1.8 4.6±2.1 Controls MMP level (ng/mL) MMP-3 24.0±14.3 MMP-12 2.0±0.5 MMP-13 4.4±0.7 Data is shown as mean ± s.d or median (IQR); n.a. (non available); *p<0.05, **p<0.01 and † p<0.1 vs. baseline. FMA (Fugl Meyer Assessement), FAC (Functional Ambulation Categories), CAHAI (Chedoke Arm and Hand Activity Inventory) and MRC (Medical Research Council). Regarding MMPs measurements, a baseline levels also at 3 and 6 months (36.2 ± 11.9 vs. 16.0 sample was obtained 10.7 ± 5.0 days after stroke ± 17.9, p<0.05 and 32.6 ± 9.5 vs. 12.9 ± 4.6, onset and no correlation was observed between p<0.05) and atrial fibrillation also with higher the MMP levels and the number of days after levels at 3 months (49 ± 6.5 vs. 28.8 ± 11.7; stroke (MMP-3: r=-0.94, p=0.761; MMP-12: p<0.05). At 6 months higher levels of MMP-12 r=0.118, p=0.688; MMP-13: r=0.296, p=0.304). and MMP-13 were detected in dyslipidemic The patients exhibited stable MMP levels during patients (2.7 ± 0.6 vs. 1.4 ± 0.4 and 6.2 ± 0.6 vs. the whole study period, and no differences with 3.0 ± 1.2, p<0.01 respectively). In the control controls (see table 3). Interestingly, MMP-12 and cohort males presented higher levels of MMP-3 MMP-13 were correlated both in controls (35.1 ± 13.6 vs. 14.1 ± 3.0, p<0.01) and higher (r=0.925, p<0.001) and in stroke patients levels of MMP-12 in subjects with diabetes (2.5 ± (r=0.951, p<0.001 at baseline; r=0.984, p<0.001 0.4 vs. 1.9 ± 0.4, p<0.05). Regarding medication, at 1 month; r=0.993, p<0.001 at 3 months and higher levels of MMP-12 were found in subjects r=0.979, p<0.001 at 6 months), but no medicated with antiplatelets, statins and of relationship was found with MMP-3. MMP-13 in those subjects taking antiplatelets, statins or antihypertensive drugs (data not Regarding MMP-3 in stroke patients, age was shown). positively associated with higher levels at three and six months (r=0.676, p<0.05 and r=0.873, High levels of MMP-12 and MMP-13 are p<0.01, respectively), hypertension with higher associated with stroke severity 74 We found a strong association between baseline were diagnosed with a TACI and presenting the MMP-12 and MMP-13 levels and stroke severity, most extensive infarcts (p=0.009 for MMP-12 and as measured by the NIHSS score on admission p=0.058 for MMP-13 at baseline; see Figure 2B). (r=0.81 and p<0.001 and r=0.67 and p=0.008, No significant association were found for MMP-3 respectively; see Figure 2A). MMP-12 was also (see Figure.2A-B), except for a statistical trend increased in patients with higher motor scores at correlating higher levels of baseline MMP-3 with the NIHSS (r=0.625 and p=0.017). Interestingly, patients with a lower motor NIHSS score on these two MMPs were elevated in patients who admission (r=-0.514 and p=0.072). Figure 2. Association of MMP levels and baseline stroke severity and extension. A) Scatter plots showing significant associations between baseline MMP-12 and MMP-13 with NIHSS scores on admission. B) Bar graphs showing higher levels of MMP-12 and MMP-13 in patients classified as TACI vs. those with other classifications (LACI, PACI, POCI), according to the OSCP classification. Elevated MMP-3 and decreased MMP-12 or outcome at three and six months in the lower MMP-13 levels in patients with a better motor proximal extremity MRC scale (p=0.026, score during rehabilitation respectively for baseline MMP-12 levels). These Despite the strong relationship between data are also represented in supplementary pre-rehabilitation and 6 months scores for Figure 1. several functional/motor scores (see The best motor and functional recovery is supplementary Table 1), higher levels of MMP-3 associated with higher plasma levels of MMP-3 were found in the patients presenting higher and lower plasma levels of MMP-12 and walking velocity at one month of IRT (r=0.579 MMP-13 and p=0.079). In addition, lower baseline levels Our results reveal that patients with higher levels of MMP-12 and MMP-13 were observed in of MMP-3 at one month presented larger patients who experienced a better outcome in improvements in motor function, assessed by the the MRC upper distal extremity scale (p=0.006 FMA score, at the same time point (r=0.632 and for MMP-12 and p=0.039 for MMP-13) at one p=0.05); this was also the case at 3 months month after or in patients who had a better (r=0.780 and p=0.013) but to a lesser extent at 6 75 months (r=0.596 and p=0.097), as shown in months (r=0.566 and p=0.07) and at the study Figure 3A. Similarly, we found that higher levels end of the study (r=0.737 and p=0.037); see of MMP-3 were associated with major Figure 3A. improvements in the CAHAI score at three Figure 3. Associations between MMP-3, MMP-12 or MMP-13 and motor recovery. A) Scatter plots showing a positive correlation between MMP-3 and improvement in FMA or CAHAI scores at different time points during the study period. B) Bar graphs showing that high MMP-3 levels one month after starting rehabilitation therapy predicted motor improvement assessed by MRC or FMA scores. C) Bar graphs showing FAC score and the 10-meter walk tests improvement in patients with lower MMP-3. D) Bar graphs showing elevated plasma MMP-12 and MMP-13 in patients that did not improve in the walking and FAC tests. During rehabilitation therapy, patients were and p=0.028, respectively) at three months and grouped into those who improved and those who for FMA scores at six months (p=0.003), as shown did not improve for each of the outcome in Figure 3B. These results suggest the predictive measures in the terms defined in the Methods value of MMP-3 for identifying patients who will section. We confirmed our findings for MMP-3 respond better to rehabilitation. because stroke patients who responded better to rehabilitation therapy presented higher plasma Two contradictory results connected levels of this protein at one month than patients improvement on the FAC score and in the who did not show improvement in the MRC of walking test at three months with lower levels of the upper proximal extremity (p=0.015) or in the MMP-3 (p=0.028 and p=0.033, respectively), as MRC of the upper distal extremity score shown in Figure 3C. In the same line of results, (p=0.014). The same results were observed for the increase/decrease in the walking velocity at FMA and MRC upper proximal scores (p=0.003 three and six months was inversely correlated 76 with baseline MMP-3 levels (r=-0.647, p=0.031 includes several types of health interventions, and r=-0.683, p=0.02; respectively). including early admission to specialized stroke rehabilitation units, intensive rehabilitation Finally, plasma MMP-12 and MMP-13 were therapies, and task specific oriented therapies (3, associated with the poorest (Figure 3D). One 5, 21) with the goal to help survivors become as month after starting the rehabilitation program, independent as possible so that they can conduct we observed the highest levels of MMP-12 and daily living activities and achieve the best MMP-13 in the patients that did not show an possible quality of life in the long-term. As we improved FAC score (p=0.012 and p=0.002, have reported for the stroke rehabilitation cohort respectively) and high MMP-12 in those that did in the present study, the most significant not show an improvement in the walking test improvements are achieved within the first one (p=0.051). Moreover, higher MMP-13 at one to three months, but recovery continues for up month was associated with the poorest level of to 6 months, as described (22). performance in the walking test at the third month (p=0.032). Although recent meta-analyses have shown strong evidence for stroke rehabilitation intervention (23, 24), daily practice indicates that individual patients respond differently to therapy, DISCUSSION which raises several questions regarding which The present study describes for the first time the patients will benefit the most and which patients temporal profile of plasma MMP-3, MMP-12 and will respond better to larger and more intensive MMP-13 up to 6 months after ischemic stroke in programs. Questions relating to the underlying patients receiving IRT. Our results describe the biological mechanisms of neural plasticity that association of MMP-12 and MMP-13 with stroke drive improvements in neurological functions severity and the poorest neurological outcomes remain to be answered, to known whether they and identify MMP-3 as a potential molecule to can be therapeutically modulated to improve monitor greatest improvements, suggesting that rehabilitation therapies. MMP-3 could serve as a biomarker of post-stroke recovery and could help in identifying those MMPs are proteolytic enzymes that can degrade patients that will respond better to the most components of the extracellular matrix and rehabilitation therapy. remodel the pericellular space. They are normally expressed in most tissues and take part Over the last decades, acute stroke care has in multiple physiological processes (25). However, improved, with the development of stroke units, uncontrolled MMP activity is involved in the hyperacute treatment with recombinant tissue pathophysiology of many disorders with plasminogen activator, and new thrombectomy underlying inflammation processes. In cerebral strategies (18-20). However, rehabilitation nervous system-related diseases, some MMPs therapies remain the only approved treatments have been associated with neurodegeneration, for the millions of patients who each year survive demyelination, neurotoxicity and blood brain a stroke but experience functional deficits. barrier dysfunction (26, 27), but emerging Evidence-based stroke rehabilitation care evidence has positioned these proteases as key 77 players in tissue regeneration processes, of transient ischemia via MMP-9 downregulation including post-stroke repair (8, 26). To our (34). Again, published studies have related brain knowledge, our study shows the longest MMP-12 expression with larger infarct size and temporal profile of MMP levels in stroke patients: worse neurological outcome, which is in from the sub-acute to the chronic phase. agreement with our results. Plasma MMP-3, MMP-12 and MMP-13 remained Finally, the most striking results of our study are stable over time without major fluctuation but related to plasma MMP-3, a member of the with two opposite behaviors. In the current study, stromelysin sub-family that degrades multiple elevated baseline MMP-12 and MMP-13 levels extracellular matrix components, activates were found in patients who had more severe growth factors and cleaves adhesion molecules, infarcts, whereas MMP-3 remained independent among other functions (26). We demonstrate, for of stroke characteristics. Previous results from the first time, an association for MMP-3 not only our group have described the expression of with better neurological status but also with nuclear MMP-13 in neuronal cells after ischemia, improved motor recovery at later time points where it is acutely elevated in the infarct tissue during rehabilitation therapy. Importantly, (28, 29). In another study, we reported the role baseline MMP-3 levels were not related to stroke of MMP-13 as a plasma biomarker for lesion characteristics nor were they correlated with expansion, as measured in diffusion-weighted baseline motor/functional prerehabilitation images in the first days following stroke (30). scores. On the other hand, when assessing With regards to its function as a major improvement during the follow-up using the 10 collagenase, other authors have shown that meter walk test we observed contradictory MMP-13 expression is increased in the results. Here we could speculate that important BBB-damaged vessels of spontaneous achievements such as better stability could not hypertensive rats (31). These data support our translate into higher walking velocity. MMP-3 finding that higher baseline MMP-13 levels are also, like other MMPs, plays a dual related to infarct extension and stroke severity. detrimental/remodeling role, and an extensive However, other investigations have reported the literature supports its central position in brain activation of MMP-13 by two weeks after tissue embryonic and postnatal development, in ischemiareperfusion, which is associated with addition to its involvement in axonal extension, perineuronal matrix remodeling (32) and synaptic plasticity and remyelination (26, 35). suggests its role in neuronal remodeling. Less is Experimental models of stroke have shown that known about MMP-12, which is also named MMP-3 expression is critical for intracranial macrophage elastase. In an experimental bleeding after tPA administration (36) and for hemorrhagic stroke model, MMP-12 expression perineuronal agrin cleavage after was associated with poor functional recovery of ischemiareperfusion in rats (37). Interesting forelimb function and with the development of results how a single administration of a broad secondary injury (33). More recently, the specific spectrum MMP inhibitor (GM6001) after knockdown of MMP-12 by the administration of ischemia, by protecting BBB integrity acutely, a shRNA-expressing plasmid (silencing the gene) enhanced vascular remodeling at three weeks resulted in brain neuroprotection in a rat model related to increased MMP-3 and tight junction 78 expression (38). determine the exact effect of rehabilitation therapy on MMP levels. However, this is for the The use of biological samples to detect the understandable reason that ethical principles presence of specific molecules with the aim of demand that all patients admitted at our hospital monitoring ongoing pathophysiological processes who are candidates for receiving IRT must be and predicting neurological outcomes for stroke enrolled in the program. In addition, the lack of survivors is the focus of current clinical and power due to the number of patients did not experimental research (39, 40). However, very allow us to perform multivariate analysis to few studies have investigated the use of explore possible interactions between the clinical biomarkers during rehabilitation therapy. Blicher variables and the biomarkers. The designing of and colleagues studied the neurotransmitter future multicenter studies will overcome this GABA by magnetic resonance spectroscopy in limitation and allow a more detailed analysis. relation to motor function after 2 weeks of constrained-induced movement therapy and showed an association between the extend of motor improvement and the magnitude of the CONCLUSIONS/IMPLICATIONS GABA/creatine ratio (41). A second study recently The ability to monitor stroke recovery using reported the use of 8-hydroxy-2’-deoxyguanosine biomarkers would be of great help when making (8-OHdG), a sensitive biomarker for oxidative decisions related to a patient’s management and DNA damage that is detectable in urine samples, rehabilitation programs. To our knowledge, this is in a cohort of stroke patients undergoing the first report on ischemic stroke patients who rehabilitation therapy (42). The authors reported were enrolled to IRT to describe the temporal a negative correlation between FMA scores and profile of plasma MMPs over a 6-month 8-OHdG levels after rehabilitation treatment and follow-up period, and we show exciting results lower levels of baseline in patients that exhibited regarding the association of MMP-3 with the best more improvement in motor function, suggesting motor recovery. This observation could be the use of 8-OHdG as a biomarker of stroke related to the role of MMPs in brain plasticity recovery. and tissue regeneration. Regardless of the Nonetheless, several caveats of this study should preliminary nature of our study and the need to be discussed. First, our study does not provide a design larger studies in a multi-center setting, we control cohort of stroke patients with similar can conclude that MMPs have emerged as characteristics that did not receive IRT to potential biomarkers of recovery during rehabilitation therapy following stroke. ACKNOWLEDGMENTS AR is supported by the Miguel Servet program of Carlos III), FM holds a fellowship from the China the Spanish Ministry of Health (Instituto de Salud Scholarship Council (CSC), AB is supported by a 79 Rio Hortega contract CM13/00265 from the Investigaciones Sanitarias (PI13/01094) and the Instituto de Salud Carlos III and IR-LL by the Spanish stroke research network INVICTUS contract CM10/00063. This work has been (RD12/0014/0005) of the Instituto de Salud supported by research grants from the Fondo de CarlosIII. REFERENCES 1. Grossman AW, Broderick JP. Advances and 2013;13:130. challenges in treatment and prevention of 11. Clinical practice guideline for the ischemic stroke. Ann Neurol. 2013;74:363-372. management of stroke. Catalan Agency for 2. Bori I, Martinez C. Conceptos fundamentales Health Information, Assessment and Quality. en rehabilitacion en el ictus:.donde, como, quien, Department of Health of Catalonia. January cuando y cuanta? Neuroreparacion y 2007. Rehabilitacion tras el ictus. Barcelona. Editorial 12. Granger CV, Dewis LS, Peters NC, et al. Stroke Marge. 2010. rehabilitation: analysis of repeated Barthel index 3. Langhorne P, Bernhardt J, Kwakkel G. Stroke measures. Arch Phys Med Rehabil. rehabilitation. Lancet. 2011;377:1693-1702. 1979;60:14-17. 4. Langhorne P, Pollock A; Stroke Unit Trialists' 13. Barreca S, Gowland C, Stratford P, et al. Collaboration. What are the components of Development of the Chedoke Arm and Hand effective stroke unit care? Age Ageing. Activity Inventory: Theoretical constructs, item 2002;31:365-371. generation, and selection. Top Stroke Rehabil.2004; 11:31-42. 5. Teasell R, Rice D, Richardson M, et al. The next revolution in stroke care. Expert Rev Neurother. 14. Arya KN, Verma R, Garg RK. Estimating the 2014;14:1307-1314. minimal clinically important difference of an upper extremity recovery measure in subacute 6. Pekna M, Pekny M, Nilsson M. Modulation of stroke patients. Top Stroke Rehabil. 2011;18 neural plasticity as a basis for stroke Suppl 1:599-610. rehabilitation. Stroke. 2012;43:2819-2828. 15. Barreca SR, Stratford PW, Lambert CL, 7. Brewer L, Horgan F, Hickey A, et al. Stroke Masters LM, Streiner DL. Test-Retest reliability, rehabilitation: recent advances and future validity, and sensitivity of the chedoke arm and therapies. QJM. 2013;106:11-25 hand activity inventory: A new measure of 8. Rosell A, Lo EH. Multiphasic roles for matrix upper-limb function for survivors of stroke. Arch metalloproteinases after stroke. Curr Opin Phys Med Rehabil. 2005;86:1616-1622. Pharmacol. 2008;8:82-89. 16. Fulk GD, Echternach JL. Test-retest reliability 9. Verslegers M, Lemmens K, Van Hove I, Moons and minimal detectable change of gait speed in L. Matrix metalloproteinase-2 and -9 as individuals undergoing rehabilitation after stroke. promising benefactors in development, plasticity J Neurol Phys Ther. 2008;32:8–13. and repair of the nervous system. Prog Neurobiol. 17. Mehrholz J, Wagner K, Rutte K, Meisner D, 2013;105:60-78. Pohl M. Predictive Validity and Responsiveness of 10. Riba-Llena I, Jarca CI, Mundet X, et al. the Functional Ambulation Category in Investigating silent strokes in hypertensives: a Hemiparetic Patients After Stroke. Arch Phys Med magnetic resonance imaging study (ISSYS): Rehabil. 2007;88:1314–1319. rationale and protocol design. BMC Neurol. 80 18. Stroke Unit Trialists’ Collaboration (2007) 29. Cuadrado E, Rosell A, Penalba A, et al. Organised inpatient (stroke unit) care for stroke. Vascular MMP-9/TIMP-2 and neuronal MMP-10 Cochrane Database Syst Rev: CD000197. up-regulation in human brain after stroke: a combined laser microdissection and protein array 19. Berkhemer OA, Fransen PS, Beumer D, et al.; study. J Proteome Res. 2009;8:3191-3197. MR CLEAN Investigators. A randomized trial of intraarterial treatment for acute ischemic stroke. 30. Rosell A, Alvarez-Sabin J, Arenillas JF, et al. A N Engl J Med 2015; 372: 11-20. matrix metalloproteinase protein array reveals a strong relation between MMP-9 and MMP-13 20. Jovin TG, Chamorro A, Cobo E, et al.; with diffusionweighted image lesion increase in REVASCAT Trial Investigators. Thrombectomy human stroke. Stroke. 2005;36:1415-1420. within 8 hours after symptom onset in ischemic stroke. N Engl J Med. 2015;372:2296-2306. 31. Ueno M, Wu B, Nishiyama A, et al. The expression of matrix metalloproteinase-13 is 21. Teasell RW, Foley NC, Salter KL, Jutai JW. A increased in vessels with blood-brain barrier blueprint for transforming stroke rehabilitation impairment in a stroke-prone hypertensive care in Canada: the case for change. Arch Phys model. Hypertens Res. 2009;32:332-338. Med Rehabil. 2008;89:575-578. 32. Nagel S, Sandy JD, Meyding-Lamade U, 22. Lee KB, Lim SH, Kim KH, et al. Six-month Schwark C, Bartsch JW, Wagner S. Focal cerebral functional recovery of stroke patients: a ischemia induces changes in both MMP-13 and multi-time-point study. Int J Rehabil Res. aggrecan around individual neurons. Brain Res. 2015;38:173-180. 2005;1056:43-50. 23. Lohse KR, Lang CE, Boyd LA. Is more better? 33. Wells JE, Biernaskie J, Szymanska A, Larsen Using metadata to explore dose-response PH, Yong VW, Corbett D. Matrix relationships in stroke rehabilitation. Stroke. metalloproteinase (MMP)-12 expression has a 2014;45:2053-2058. negative impact on sensorimotor function 24. Veerbeek JM, van Wegen E, van Peppen R, et following intracerebral haemorrhage in mice. Eur al. What is the evidence for physical therapy J Neurosci. 2005;21:187-196. poststroke? A systematic review and 34. Chelluboina B, Warhekar A, Dillard M, et al. meta-analysis. PLoS One. 2014;9:e87987. Post-transcriptional inactivation of matrix 25. Nagase H, Visse R, Murphy G. Structure and metalloproteinase-12 after focal cerebral function of matrix metalloproteinases and TIMPs. ischemia attenuates brain damage. Sci Rep. Cardiovasc Res. 2006;69:562-573. 2015;5:9504. 26. Van Hove I, Lemmens K, Van de Velde S, 35. Van Hove I, Verslegers M, Buyens T, et al. An Verslegers M, Moons L. Matrix aberrant cerebellar development in mice lacking metalloproteinase-3 in the central nervous matrix metalloproteinase-3. Mol Neurobiol. system: a look on the bright side. J Neurochem. 2012;45:17-29. 2012;123:203-216. 36. Suzuki Y, Nagai N, Umemura K, Collen D, 27. Rosenberg GA. Matrix metalloproteinases Lijnen HR. Stromelysin-1 (MMP-3) is critical for and their multiple roles in neurodegenerative intracranial bleeding after t-PA treatment of diseases. Lancet Neurol. 2009;8:205-216. stroke in mice. J Thromb Haemost. 28. Cuadrado E, Rosell A, Borrell-Pages M, et al. 2007;5:1732-1739. Matrix metalloproteinase-13 is activated and is 37. Sole S, Petegnief V, Gorina R, Chamorro A, found in the nucleus of neural cells after cerebral Planas AM. Activation of matrix ischemia. J Cereb Blood Flow Metab. metalloproteinase-3 and agrin cleavage in 2009;29:398-410. 81 cerebral ischemia/reperfusion. J Neuropathol Exp comprehensive systematic review. Br J Clin Neurol. 2004;63:338-349. Pharmacol. 2012;74:230-240. 38. Yang Y, Thompson JF, Taheri S, et al. Early 41. Blicher JU, Near J, Nass-Schmidt E, et al. inhibition of MMP activity in ischemic rat brain GABA levels are decreased after stroke and GABA promotes expression of tight junction proteins changes during rehabilitation correlate with and angiogenesis during recovery. J Cereb Blood motor improvement. Neurorehabil Neural Repair. Flow Metab. 2013;33:1104-1114. 2015;29:278-286. 39. Cucchiara B, Montaner J. Blood biomarkers 42. Hsieh YW, Lin KC, Korivi M, Lee TH, Wu CY, Wu for predicting outcome after stroke: reading the KY. The reliability and predictive ability of a tea leaves. Neurology. 2013;80:1270-1271. biomarker of oxidative DNA damage on 40. T Hasan N, McColgan P, Bentley P, Edwards RJ, functional outcomes after stroke rehabilitation. Sharma P. Towards the identification of blood Int J Mol Sci. 2014;15:6504-6516. biomarkers for acute stroke in humans: a 82 Supplementary Table1. Association study between pre-rehabilitation and 6 months scores of functional and motor measures. RANKIN FMA CAHAI WALKING SS-QOL MRC (Upp-Prox) MRC (Upp-Dis) MRC( Inf-Prox) MRC (Upp-Dis) r r r r r 0-4 5 0-4 5 0-4 5 0-4 5 NIHSS -0.55† 14.5 (12-17) 7 (4.5-8.5)* (admission) NIHSS motor 7.5±0.7 3.0±2.6* (admission) NIHSS 0.723** -0.678* -0.568† 10.5 (8.3-13.5) 5 (4-7)* 10.5 (8.3-13.5) 5 (4-7)* 13 (12-14) 5 (4.5-8.5)* 12 (8-14) 5 (4.3-8.5)* (3-4days) NIHSS motor 0.586* -0.797** -0.67* -0.597† -0.812** 5.5±2.6 2.1±2.2* 5.5±2.6 2.1±2.2* 7.5±0.7 2.4±2.1* 6.7±1.5 2.1±2.0** (3-4days) RANKIN 0.835** -0.681* -0.575† -0.681* -0.592† 4 (4-4) 3 (3-4)* 4 (4-4) 3 (3-4)* 4 (4-4) 3 (3-4)† BI -0.871** 0.888** 0.745** 0.603* 0.670* 36 (18.3-41) 79 (62-95)* 36 (18.3-41) 79 (62-95)* 22.5 (14-31) 70 (41-89)† 31 (14-41) 74.5 (46.3-92)* FMA -0.707* 0.841** 0.784** 0.528† 0.602† 4 (2.5-6.3) 50 (23-55)* 4 (2.5-6.3) 50 (23-55)* 3 (2-40) 47 (6.5-54)* 4 (2-4) 48.5 (11-54.5)* FAC -0.836** 0.789** 0.674* 0.765** 0.696* 0 (0-0.8) 3 (1-5)* 0 (0-0.8) 3 (1-5)* 0 (0-0) 2 (0.5-4.5)† 0 (0-0) 2.5 (1-4.8)* CAHAI -0.612* 0.875** 0.876** 13 (13-13) 70 (13-71)* 13 (13-13) 70 (13-71)* 13 (13-13) 68.5 (13-71)† WALKING -0.791** 0.711* 0.58† 0.813** 0.879** 0 (0-0.4) 1 (0.5-1)* 0 (0-0.4) 1 (0.5-1)* 0 (0-0) 0.5 (0.3-1)† 0 (0-0) 0.8 (0.5-1)* MRC -0.536† 0.860** 0.879** 0 (0-0) 4 (4-4)** 0 (0-0) 4 (4-4)** 0 (0-0) 4 (4-4)* (Upp-Prox) MRC 0.690** 0.593† (Upp-Dis) MRC -0.566† 0.609* 0.548† 0.673* 1 (0.3-3.3) 4 (4-5)* 1 (0.3-3.3) 4 (4-5)* 0 (0-1) 4 (4-5)* (Inf-Prox) MRC 0.571† 0.588† 0.732* 0 (0-0) 4 (2-5)† 0 (0-0) 4.5 (4-5)* (Inf-Dis) Note: ** p <0.01, *p <0.05 and † p<0.1 Supplementary Figure 1. MMP levels and motor scores during rehabilitation. A) Correlation plot showing the association between MMP-3 levels and the walking velocity assessed at 1 month. B) Bar graphs showing the difference in baseline MMP-12 and MMP-13 levels between patients that showed normal or weakened muscle strength at different time points assessed by the MRC score. 85 86 Discussion 4. Discussion 4.1. MMP-13 deficiency protects mice from ischemia-reperfusion injury whereas high level of plasma MMP-12 or -13 is associated with stroke severity in ischemic stroke patients. The present study firstly demonstrates that MMP-13 participates in acute damage which is enhanced during the reperfusion phase in a mouse model of stroke and related to the presence of spontaneous HT. In parallel we have described for the first time the association of MMP-12 and MMP-13 with stroke severity and with the poorest neurological/motor outcomes in patients under neurorehabilitation therapy. It is known that ischemic damage induces the unregulated expression of several MMPs at multiple levels (brain and plasma), which has been documented previously both in humans and rodent models of cerebral ischemia71,123,124,125. Most of these papers focus on MMP-9, a gelatinase that has been extensively studied in the ischemia context, by proving that its gene or pharmacological inhibition protects from ischemia injury and reduces hemorrhagic transformations123,126,127. However we know that other MMPs are also dysregulated in the human infarct tissue suggesting its implication in tissue damage. In this regard a previous study from our group showed that several members of the MMP-family of proteases, including MMP-9 and -13 among others, were upregulated in the post-mortem infarct tissue of patients who had an ischemic stroke within the previous 4 days when compared to the contralateral hemisphere81. Furthermore, it is known that MMP-9 is able to be activated directly after stroke onset, with the plasma level correlated with the infarct size and functional outcome87,128 , while only a few studies shows the MMP-13 activation in the brain/plasma related to lesion expansion in the acute phase 101,102. More recently, increased MMP-13 mRNA and protein levels were observed in brain tissue during the acute phase in a rodent model of ischemia-reperfusion129. The two primary outcomes in preclinical studies are infarct lesion reduction and the improvement in neurological function which might translate to an improved neuroprotection or functional outcome clinically. The quantification of infarct size can be obtained by several methods such as TTC staining for experimental models or MRI for either animal or human subjects. However, more variability exists on the tests used to assess functional status or neurological outcomes in 130,131,132,133,134 rodents , whereas in humans, the NIH stroke Scale score and the modified Rankin scale being the gold standard methods to evaluate stroke severity and functional outcome135,136,137, respectively. 87 Our results show a significant reduction on infarct size in MMP-13 gene knock-out (KO) mice compared to wild-type (WT) at one and three days after cerebral ischemia-reperfusion, but not after permanent MCAo although infarct volumes were also smaller in MMP-13 deficient mice. In fact, differences between the two models were observed in the WT groups, being infarcts of the ischemia-reperfusion model larger than in the permanent occlusion but not significantly. The observed neuroprotection in terms of infarct volume was correlated with our observation on the reduced incidence of spontaneous HT events in MMP-13 KO mice. Spontaneous HT occurs in patients with ischemic stroke mainly between the 3th and the 7th day of stroke onset138, and the disruption of BBB is a central step that precedes HT development. Early within the first 2 hours, the opening of BBB has been observed in stroke patients and enhanced in later phases139. In the delayed HT, the opening of BBB is thought to be mediated by MMPs, other brain proteases, neuroinflammation, as well as vascular remodeling and neovascularization 30. It has been described that MMPs derived from blood have a dominant role in the early BBB disruption and HT, while the brain derived MMPs such as MMP-9, MMP-2 and MMP-3 contribute to the delayed HT140,141,142. Related to endogenous neurorepair, we have explained that vascular remodeling and angiogenesis are important for the recovery after stroke143, and pro-angiogenic factors may also promote post-stroke HT144,145. For example, VEGF is an important factor involved in angiogenesis and a decrease on VEGF reduces HT in rodent stroke model146. Our results show a tendency to an increased incidence of hemorrhage at day 3 which might indicate the involvement of MMP-13 in the delayed HT. Although HT is associated with the size of ischemic area147, our study seems to be independent to this fact since similar infarct size were observed at 24 hours and 72 hours after ischemia-reperfusion in WT mice who increased the incidence of HT at 72 hours. At the same time, no such increase was found in MMP-13 KO mice, indicating that MMP-13 deficiency may protect the mice brain tissue not only from initial tissue injury but also from the upcoming hemorrhage. Regardless the day of euthanasis, WT mice presented increased incidence of HT events than MMP-13 KO mice. We have also observed the decreased amounts of brain angiogenic factors including VEGF and Ang-2 in MMP-13 KO mice at day 7 after ischemia. The combination of VEGF and Ang-2 has been demonstrated to lead angiogenesis related to BBB disruption by increasing the activity of MMP-9 after 2 weeks treatment in mice148. The biphasic role of VEGF in stroke has been described in a previous study with early VEGF promoting HT and later VEGF promoting BBB integrity and vessel function149. 88 Discussion With this background, MMP-13 may also play a similar dual role directly and indirectly in HT in the first 7 days after stroke. One unexpected finding is the presence of striatal infarctions at day 3 after ischemia-reperfusion, related to the presence of hemorrhage within the infarct area. Since our model or MCA occlusion only affects the cortical territory, we hypothesize that the growing formation of a hematoma induced by the hemorrhage may constrict the vascular supporting blood flow to striatum. Unfortunately, most studies based on neuroprotection assess infarct size and extension at 24-48 hours making difficult to compare this finding with others. However, our study also includes a longer time follow-up of the mice to study neurorepair mechanisms, and it seems that the final functional outcome at day 14 was not different between MMP-13 KO and WT mice which presented larger infarcts and more incidence of HT events acutely (although these observations are performed in separate study groups). This could suggest that WT mice have repaired/recovered better or faster than that in MMP-13 deficient mice and our next findings support this hypothesis. The other primary outcome mentioned at the beginning of this section is the functional outcome. In this regard, the appropriate tests should be chosen accurately according to the stroke models, evaluation time and animal species even strains130. The grip strength test is highly specific, well-defined aspect of behavior and less stressful to the animals than other tests. KIlic et al. found that grip test is sensitive for detecting outcomes from different treatments in C57BL/6 mice after mild transient ischemia at day 7 as well as day 30150. In our hands, a significantly reduced forelimb force was also found from day 1 until day 14 in permanent distal MCA occluded Balb/c mice130. In the present study, the grip strength test has proven to detect reduced forelimb function after transient MCAo in C57BL/6 mice as long as 14 days after ischemia. At day 1, all mice decreased the forelimb force however this decrease was only significant in WT mice, in agreement with our findings on infarct volume and HT incidence. During the follow-up only the decreased forelimb force in WT versus MMP-13 KO was maintained at one week but not at two weeks. This fact could indicate different repair mechanisms between genotypes. It is interesting to link the MMP-13 observation in the animal model of ischemia with the findings in patients under rehabilitation therapy after stroke: high levels of MMP-13 were observed in those patients with more extensive infarcts (TACI classification) and presenting a more severe neurological status at baseline 89 (measured by the NIHSS). Although blood samples were obtained in some cases days after the baseline neurological evaluation no association was found between plasma MMP levels and the day of baseline blood sampling. The expression of MMP-13 is increased in the BBB-damaged vessels of rats151, and the role of plasma MMP-13 as a biomarker for the lesion expansion has been proposed in a previous study from our group102. Together with our finding in the experimental model, we can conclude that MMP-13 contributes to the acute cerebral damage further contributing in BBB damage. It would have been interesting to analyze the association of MMP-13 level with the presence of spontaneous HT, however the sample size was too small to speculate on this interesting point that certainly deserves to be further investigated. Overall, we demonstrate the role of MMP-13 in brain injury after ischemia however our group and other have reported the active role of other MMPs related to brain infarcts, as detailed in the introduction section. Therefore, which MMP is more important? Are these MMP actions related? The activation of MMP proteins by other members of the family has been widely described152,153. We have studied the protein levels of infarct MMP-3, -8 and MMP-9 compared to their contralateral hemisphere describing a significant increase at day 3 after ischemia in both genotypes, with a significant decrease of all studied MMPs in the ipsilateral hemisphere of the MMP-13 KO mice compared to WTs. In our study, the detection of MMPs was designed to measure the total amount of MMPs including pro and activated forms except for MMP-9 (only the pro-form) due to manufacturer setting of the kit. We included one collagenase (MMP-8), one relevant gelatinase as MMP-9 and one stromelysin (MMP-3) as representative proteins of the MMP family that could undergo compensatory, up- or down- regulations in the knockdown model of MMP-13. Our conclusion is that the decreased level of MMP-3, -8 and pro-MMP-9 in MMP-13 KO mice in the context of cerebral ischemia indicates that the interaction of MMP-13 with other members of MMPs family might be controlling the MMP cascade. MMP-8 is known as the “neutrophil collagenase” targeting primary fibrillar collagen which is upregulated in the infarct tisue81. In our study, the knockdown of MMP-13 does not stimulate the overexpression of MMP-8, instead inhibits the expression of this collagenase after ischemia indicating the possible central and upstream role of MMP-13. In the present study, we may conclude that MMP-13 deficiency affects the expression of MMP-9, and this gelatinase has been directly involved in both injury and repair mechanisms after stroke71,76. In a previous study from our group both MMP-9 and MMP-13 showed an association with infarct lesion expansion after 90 Discussion stroke in a multiple screening protein array of the MMP family102. Other studies have demonstrated that the knockdown of MMP-12 inhibited the increased MMP-9 activity in ischemic brain154 suggests that the interaction among MMPs in the ischemia context is possible. 91 4.2. MMP-13 deficiency inhibits spontaneous neurorepair by reducing the number of proliferating neuroblasts and angiogenesis in peri-infarct areas, while decreased level of plasma MMP-12 or -13 is associated with better motor recovery in patients. This thesis shows how the deficiency in MMP-13 reduces the amount of newly born neuroblasts in the peri-infarct areas during neurological recovery and alters stroke-induced angiogenesis by reducing peri-infarct vessel density and suppressing the amounts of G-CSF, Ang-2 and VEGF-A after cerebral ischemia. In stroke patients, high MMP-12 and -13 levels during rehabilitation were found in patients with a worst recovery, however those levels were found associated to stroke severity at baseline. We have learnt that the regulation of MMPs during post-stroke might promote a therapeutic effect71 at both acute and stroke repair phases, and most therapeutic approaches under investigation now focus on the enhancement of endogenous repair mechanisms such as neurogenesis and angiogenesis155. In our experimental ischemia study, a similar cortical atrophy and functional outcome were observed among genotypes at day 14 after ischemia, which could be a balanced consequence of both the inhibition of cell apoptosis and/or promotion of post-stroke repair mechanism. We prove that as least angio-neurogensis progression is altered in mice lacking MMP-13. Our results show a significant reduction in the number of proliferating neuroblasts (BrdU+/DCX+ cells) migrated from the SVZ towards to the peri-infarct cortex in MMP-13 KO mice at day 14 after stroke, despite the total amount of neuroblasts in those areas did not change among genotypes. We explored if new mature neurons were generated in those areas but numbers were similar among genotypes. In this regard, a study focused on toll-like receptor 4 in focal cerebral ischemia model described some BrdU labeled cells in the peri-infarct cortex co-stained with an interneuron marker at day 14, indicating that the new generated neuroblasts after stroke were able to maturate into neurons156. In the same line, Saha et al. found that, only a few number of proliferating cells present in areas surrounding the lesion expressed NeuN, while most of them co-express GFAP and Olig2157. Therefore, it seems possible that other neuronal cells or progenitors could be under proliferation in our peri-infarct tissue. Importantly, a positive association between the number of proliferating neuroblast and vessel density was observed in the present study. Neurogenesis has been shown to be coupled to angiogenesis in a post-stroke neurovascular niche158. Other authors have also identified that the proliferating neuroblasts in peri-infarct areas are closely 92 Discussion associated to the blood vessels159. The mechanism underlying this crosstalk may involve the vascular network acting as migration scaffolds as well as being the source of pro-neurogenic and migratory factors such as VEGF and SDF-1160. Delayed inhibition of SDF-1 at day 7 suppresses neovascular remodeling and impairs functional recovery at day 1493. In line of these studies we have observed smaller vessel density in ipsilateral peri-infarct tissue of MMP-13 KO mice compared to WTs at 14 days, but not at day 7 post-ischemia. However, we found a suppressed amount of thropic factors in MMP-13 KO mice at day 7, such as VEGF-A, G-CSF and Ang-2, which occurs prior to the observation of decreased vascular density. Additionally, these factors may contribute to the recovery after ischemia damage: VEGF-A is one of the VEGF family, which plays multiple roles in the normal adult brain, including vascular homeostasis, neurogenesis, synaptic plasticity and neuronal activity161 and similar functions have been found after pathological stimuli162. VEGF-A protein expression increases in neurons, astrocytes and macrophages after ischemia in rats163. At 24 hours after MCAo in rats, intraventricular administration of VEGF-A for 3 days increased the post-ischemic BrdU labeling in neuronal cells as well as immunoreactive endothelial cells 164. What’s more, the effect of a cell-based therapy aiming to enhance post-stroke functional recovery could be enhanced by the combination with VEGF-A in several MCAo models162. G-CSF is a hematopoietic growth factor produced by activated monocytes as well as endothelial cells, fibroblasts, mesothelial cells and platelets165,33. G-CSF and its receptors are expressed in response to ischemia in the penumbral region of ischemic stroke, and markedly improve long-term behavioral outcome166. The up-regulated expression in the subacute phase has been proven predominantly in vascular cells167 and the G-CSF administration after tMCAo enhances angiogenesis as well as neurogenesis at day 7 after ischemic brain injury in rats168. Additionally, Ang-2 is typically expressed at sites of vascular remodeling in the adult tissue, which is dramatically increased in the infarct cortex and maintained for as long as 7 days after ischemia in mice169,170. The combination of VEGF and Ang-2 promotes more angiogenesis at 2 weeks after treatment in the adult mice brain compared to VEGF alone148. Besides, Ang-2-enhanced migration and neuronal differentiation of SVZ neural progenitor cells at day 7 after MCAo in mice; while Ang-2 enhanced migration could be blocked by matrix metalloproteinase inhibitor171. Finally, EPCs involved in vascular remodeling after ischemia have been proven to secrete a combination of these factors including fibroblast growth factor (FGF-b), platelet-derived growth factor (PDGF-bb) and VEGF, among others122. FGF has shown to reduce the infarct volume in ischemic rats in the subacute phase, which may 93 involves post-stroke angiogenesis and neurogenesis172,173. The expression of PDGF is upregulated in neurons, reactive astroglial cells and pericyte after ischemic stroke in humans and animals174,175, and the platelet microparticles containing VEGF, FGF and PDGF significantly promote neural stem cell proliferation, survival and differentiation, which can be partly blocked by any specific inhibitor176. Whether the decreased neurogenesis in MMP-13 KO mice was finally caused by the direct actions of MMP-13 or by the relatively smaller infarct size was not fully addressed in the present study. A strong negative correlation between cell proliferation at the ipsilateral SVZ at day 14 and the infarct size at day 2 has been reported in a filament induced permanent MCAo model156, while other studies have demonstrated that MMP-9 directly contributes to the increased neuroblast migration after stroke, with MMP-3 being involved in the neural progenitor cells’ migration94,89. In this line we have shown results on suppressed expression of MMP-3 and -9 in the MMP-13 KO genotype which could explain a modulation on neuroblast migration. However, in stroke patients, we have observed an association between decreased level of plasma of MMP-12 and -13 with better motor recovery during rehabilitation assessed with the MRC and FAC scores. This is possibly related to the relatively high baseline MMP-12/-13 levels in patients with more severe and extensive strokes. The statistical adjustment for baseline stroke severity and other factors such as age may solve this issue, but as explained larger studies increasing the sample size are needed. 94 Discussion 4.3. MMP-13 gene silencing impairs the angio-vasculogenic function of Endothelial Progenitor Cells which shape simpler and less connected vascular networks in vitro. As discussed we have demonstrated that MMP-13 deficiency impairs spontaneous vessel formation in vivo in the peri-infarct cortex at day 14 after cerebral ischemia in mice. In the context of tissue repair, this fact could depend on the capability of endogenous EPCs actions during vessel remodeling. In this regard, we also show that silencing the MMP-13 gene impairs the EPCs’ function on tubulogenesis in Matrigel assays in vitro, indicating the key role of this protease in vessel formation. Nowadays more evidences show that EPCs are present after ischemic stroke since the first study of EPCs mobilization in response to tissue ischemia177. The maintenance of endothelial integrity plays a critical role in vascular-related diseases and EPCs have been shown to be incorporated into neovessels63,178. Taguchi et al. firstly reported that CD34þ/CD133þ cells, as a EPCs-enriched population, provided a marker of cerebrovascular function179. And several studies have proven that endogenous EPCs participate in the neovascularization via CXCR-4/SDF-1 axis after permanent MCAo in rats180. After stroke injury, it is known that there is a modulation on circulating EPCs level which produce a variety of growth factors as well as cytokines or chemokines including FGF, VEGF, PDGF-bb, CXCL4, CXCL7 or several MMPs122,181,182,183, these factors might stimulate local endothelial cell mobility, growth and fucntion 112 and other repair mechanisms such as neurogenesis or oligodendrogenesis as discussed in the EPCs’ review publication155. In this regard the first observations after EPCs identification suggested that these cells contributed to neovascularization by differentiating into cells forming structural components of the vasculature 63,177,184. However the paracrine hypothesis that EPCs contribute to vascular growth primarily by secreting proteins and recruiting additional progenitor/stem cells to the injury site has become more acceptable185,186,187. MMPs play a key role in the degradation of basal membrane, one of the earliest steps of new vessel formation during angiogenesis or vasculogenesis. MMP-13 acts as a collagenase by cleaving the interstitial collagens (I, II, and III) and has a key role in the MMP activation cascade188. The angio-vasculogenic response of EPCs has been reported to be impaired in MMP-9 deficient mice in an MCAo model95,126 and in a hindlimb ischemia model189. In this context, we demonstrate for the first time the role of MMP-13 on EPCs function in an in vitro model of tube formation. The outgrowth endothelial cells (OECs) used in the present study are a homogenous and 95 highly proliferating cell population of endothelial-like cells which directly participate in tubulogenesis190, while endothelial cells proliferation can be stimulated by the EPCs secretome191. Thus, the final in vivo influence of MMP-13 on endogenous or therapeutic EPCs function will influence not only the direct impairment of vessel formation, but also to the secretion of angiogenic/trophic factors. However, whether the in vitro tubulogenesis findings could translate into the in vivo setting of vessel formation stimulated by EPCs is still unknown. 96 Discussion 4.4. Elevated plasma MMP-3 is associated to motor and functional improvement during rehabilitation therapy after stroke. We have described for the first time the temporal profile of several MMPs over different time points in patients under an intensive neurorehabilitation program up to 6 months identifying MMP-3 as a potential monitoring molecule to detect those patients that will respond better to rehabilitation therapy. The use of biological samples to detect the presence of specific molecules with the aim of monitoring ongoing pathophysiological processes and predicting neurological outcomes for stroke survivors is the focus of current clinical and experimental research192,193. However, very few studies have investigated the use of biomarkers during rehabilitation therapy as explained in the introduction section and none has explored the use of MMPs. Why studying a biomarker for recovery in the context or neurorehabilitation? The daily practice for physiatrists indicates that individual patients respond differently to therapy, which raises several questions regarding which patients will benefit the most from the proposed program or which patients will respond better to larger and more intensive programs. Moreover, it is accepted that rehabilitation enhances neural/brain plasticity but no pharmacological treatments are available to modulate or enhance this plasticity111. With this aim it is imperative to identify the underlying biological mechanisms of neural plasticity that drive improvements in neurological functions. To know the temporal profile of a given molecule (such as MMPs) in circulating blood could provide us an amazing chance to develop its use as a biomarker to diagnose or predict the patient outcome. For example, plasma MMP-9 level within acute (<24 - 48 hours) phase of ischemic stroke has proved to correlate with severe outcome, large infarcts and has been considered as a biomarker for acute hemorrhage after ischemic stroke194,195,196. However, several promising markers have not finally shown sufficient evidence to support the use of any of them in the clinical practice 197 for several reasons (the source of plasma MMPs may differ from that of MMPs in brain, a single marker may not be able to reflect the complexity of a neurorepair process, etc.). MMP-3, also known stromelysin-1 presents a broad substrate specificity for extracellular matrix proteins including fibronectin, collagen, lamina and proteoglycans, as well as for other proteases including pro-MMP-3, -8, -9 and -13198,199. Neuroscience literature supports its central position in brain tissue at 97 embryonic and postnatal development, in addition to its involvement in axonal extension, synaptic plasticity and remyelination199,200. Moreover, experimental models of stroke have shown that MMP-3 expression is critical for intracranial bleeding after tPA administration201 and for peri-neuronal agrin cleavage after ischemia-reperfusion injury in rats202 while the single administration of a broad spectrum MMP inhibitor (GM6001) after ischemia protected BBB integrity acutely and enhanced vascular remodeling related to increased tight junction expression203. Our study on MMPs as potential markers of functional response to rehabilitation therapy describes an association of MMP-3, for the first time, with better neurological status at one month but also with improved motor recovery at later time points during rehabilitation therapy. Importantly, MMP-3 baseline level was not associated to stroke severity measured by NIHSS, extensive infarcts, age or other factors that could compromise the final outcome of the patient. Although the reduced number of patients studied did not allow us to perform multivariate analysis to explore possible interactions between the clinical variables and the proposed biomarkers. We have found similar levels between patients and controls and no change over time in MMP-3 levels, however those patients that showed better improvements in the FMA, MRC and CAHAI scores presented higher levels of MMP-3. Those tests are widely used and accepted to monitor motor/functional status and progression in rehabilitation programs204. In this line another publication has shown the potential use of 8-OHdG (a promising biomarker of oxidative stress) as a valid predictor of functional outcomes in patients under rehabilitation after stroke by using the FMA and MRC scores205. 98 Discussion 4.5. Future perspectives Our results show the protective actions of MMP-13 deficiency in terms of infarct volume and hemorrhagic transformation which deserves further confirmatory studies such as a the specific pharmacologic in vivo inhibition of MMP-13 together with potential recovery studies in MMP-13 KO mice by administrating recombinant MMP-13. Also, confirmatory studies in other MCAo models will provide further evidences of our findings. The in vitro tubulogenic remodeling study has demonstrated the promising role of MMP-13 in angiogenesis, but how will it modulated EPCs performance in vivo is still unknown. Since EPCs are being proposed as advanced therapies for stroke repair, new research is needed to relate the current in vitro observations to the in vivo context of cerebral ischemia in a putative EPC treatment. Finally since MMP-13 is known to activate other MMPs such as MMP-2 and MMP-9 and we have demonstrated the important role of MMP-9 in EPCS function in other studies, it could be possible that the actions described for MMP-13 are up-stream in a proteolitic cascade. Additional studies investigating the consequences of MMP-13 activation/inhibition within the MMP family would provide valuable information to identify future therapeutic targets. A multicentre study is needed to confirm our findings on MMP-3 as potential marker to monitor motor progression during rehabilitation therapy, which potentially could be used to adjust and adapt rehabilitation programs for stroke patients. Finally to better translate our results into clinical studies, the use of aged MCAo models, studies in female rodents or in mice carrying stroke co-morbidities are needed since functional outcome, neuroblast migration, vascular remodeling and the brain plasticity response might be influenced by those factors. 99 100 Conclusion 5. Conclusions The main conclusions of this PhD thesis are: 1) EPCs secretome contribute to post-stroke re-vasculatrization by mediating cell-cell communication actions. These interactions connect angiogenesis but also neurogenesis and oligodendrogenesis within a dynamic neurorepair niche. 2) MMP-13 is a crucial protease involved in both the development of acute infarct lesions after cerebral ischemia as well as vascular remodeling and neurogenesis in late phases by controlling the protein expression of other brain MMPs and trophic factors such as G-CSF, Ang-2 and VEGF-A. High levels of plasma MMP-13 as well as MMP-12 are associated with stroke severity in patients. 3) We firstly report the temporal profile of plasma MMP-3, MMP-12 and MMP-13 during rehabilitation therapy and demonstrate the potential use of MMP-3 as a biomarker of motor improvement. 101 102 References 6. References 1. Díaz-Guzmán, J. et al. Stroke and transient ischemic attack incidence rate in Spain: The IBERICTUS study. Cerebrovascular Diseases 34, 272–281 (2012). 2. Lo, E. H., Dalkara, T. & Moskowitz, M. A. Mechanisms, challenges and opportunities in stroke. Nature reviews. Neuroscience 4, 399–415 (2003). 3. Eliasziw, M., Kennedy, J., Hill, M. D., Buchan, A. M. & Barnett, H. J. M. Early risk of stroke after a transient ischemic attack in patients with internal carotid artery disease. Cmaj 170, 1105–1109 (2004). 4. Correia, M., Fonseca, A. C. & Canhão, P. Short-term outcome of patients with possible transient ischemic attacks: a prospective study. BMC Neurology 15, 1–6 (2015). 5. Beauchamp, N. J. & Bryan, R. N. Acute Cerebral lschemic Infarction:A Pathophysiologic Reviewand RadiologicPerspective. Ajr 171, 73–84 (1998). 6. Lo, E. H., Moskowitz, M. a. & Jacobs, T. P. Exciting, radical, suicidal: How brain cells die after stroke. Stroke 36, 189–192 (2005). 7. Dirnagl, U., Iadecola, C. & Moskowitz, M. a. Pathobiology of ischaemic stroke: an integrated view. 4441. Trends in Neurosciences 22, 391–397 (1999). 8. Halliwell, B. Reactive oxygen species and the central nervous system. Journal of Neurochemistry 59, 1609–1623 (1992). 9. Warner, D. S., Sheng, H. & Batinić-Haberle, I. Oxidants, antioxidants and the ischemic brain. The Journal of experimental biology 207, 3221–3231 (2004). 10. Ji, H. H., Sang, W. H. & Seung, K. L. Free radicals as triggers of brain edema formation after stroke. Free Radical Biology and Medicine 39, 51–70 (2005). 11. Lee, J. M., Grabb, M. C., Zipfel, G. J. & Choi, D. W. Brain tissue responses to ischemia. The Journal of clinical investigation 106, 723–31 (2000). 12. Astrup, S. and S. Thresholds in Cerebral Ischemia - The Ischemic Penumbra. A Journal of Cerebral Circulation 12, (1981). 13. MacManus, J. P. & Buchan, A. M. Apoptosis after experimental stroke: fact or fashion? Journal of neurotrauma 17, 899–914 (2000). 103 14. Dohmen, C. et al. Spreading depolarizations occur in human ischemic stroke with high incidence. Annals of Neurology 63, 720–728 (2008). 15. Baron, A. E. H. and J.-C. The ischemic penumbra: still the target for stroke therapies? Frontiers in Neurology 6, (2015). 16. Lees, K. R. et al. Time to treatment with intravenous alteplase and outcome in stroke: an updated pooled analysis of ECASS, ATLANTIS, NINDS, and EPITHET trials. The Lancet 375, 1695–1703 (2010). 17. Grotta, J. C. & Hacke, W. Stroke Neurologist’s Perspective on the New Endovascular Trials. Stroke 1–7 (2015). doi:10.1161/STROKEAHA.115.008384 18. T.G. Jovin, A. Chamorro, E. Cobo, M.A. de Miquel, C.A. Molina, A. R., L. San Román, J. Serena, S. Abilleira, M. Ribó, M. Millán, X. Urra, P. Cardona, E. López-Cancio, A. Tomasello, C. Castaño, J. Blasco, L. Aja, L. D. & H. Quesada, M. Rubiera, M. Hernández-Pérez, M. Goyal, A.M. Demchuk, R. von Kummer, M. Gallofré, and A. Dávalos, for the R. T. I. Thrombectomy within 8 Hours after Symptom Onset in Ischemic Stroke. 1–11 (2015). doi:10.1056/NEJMoa1503780 19. Berkhemer OA, Fransen PS, Beumer D, van den Berg LA, Lingsma HF, Yoo AJ, et al; M. C. I. A Randomized Trial of Intraarterial Treatment for Acute Ischemic Stroke. New England Journal of Medicine 372, 11–20 (2015). 20. Jin, R., Yang, G. & Li, G. Inflammatory mechanisms in ischemic stroke: role of inflammatory cells. Journal of leukocyte biology 87, 779–789 (2010). 21. Hayakawa, K., Qiu, J. & Lo, E. H. Biphasic actions of HMGB1 signaling in inflammation and recovery after stroke. Annals of the New York Academy of Sciences 1207, 50–57 (2010). 22. Tobin, M. K. et al. Neurogenesis and inflammation after ischemic stroke: what is known and where we go from here. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 34, 1573–84 (2014). 23. Spite, M. & Serhan, C. N. Novel lipid mediators promote resolution of acute inflammation: impact of aspirin and statins. Circulation research 107, 1170–84 (2010). 24. Nathan, C. & Ding, A. Nonresolving Inflammation. Cell 140, 871–882 (2010). 104 References 25. García-Yébenes, I. et al. A mouse model of hemorrhagic transformation by delayed tissue plasminogen activator administration after in situ thromboembolic stroke. Stroke 42, 196–203 (2011). 26. Eltzschig, H. K. & Eckle, T. Ischemia and reperfusion—from mechanism to translation. Nature Medicine 17, 1391–1401 (2011). 27. Abbott, N. J., Patabendige, A. A. K., Dolman, D. E. M., Yusof, S. R. & Begley, D. J. Structure and function of the blood–brain barrier. Neurobiology of Disease 37, 13–25 (2010). 28. Jin, X. et al. Spatiotemporal evolution of blood brain barrier damage and tissue infarction within the first 3h after ischemia onset. Neurobiology of Disease 48, 309–316 (2012). 29. Yang, G. Y. & Betz, A. L. Reperfusion-induced injury to the blood-brain barrier after middle cerebral artery occlusion in rats. Stroke 25, 1658–1664 (1994). 30. Jickling, G. C. et al. Hemorrhagic transformation after ischemic stroke in animals and humans. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 34, 185–99 (2014). 31. Zhang, J., Yang, Y., Sun, H. & Xing, Y. Hemorrhagic transformation after cerebral infarction: current concepts and challenges. Annals of translational medicine 2, 81 (2014). 32. Widgerow, A. D. Ischemia-reperfusion injury: influencing the microcirculatory and cellular environment. Annals of plastic surgery 72, 253–60 (2014). 33. J, M. Fisiopatología de la Isquemia Cerebral: Modelos in vivo e in vitro de isquemia cerebral. (2008). 34. Longa, E. Z., Weinstein, P. R., Carlson, S. & Cummins, R. Reversible middle cerebral artery occlusion without craniectomy in rats. Stroke; a journal of cerebral circulation 20, 84–91 (1989). 35. Lipsanen, A. & Jolkkonen, J. Experimental approaches to study functional recovery following cerebral ischemia. Cellular and Molecular Life Sciences 68, 3007–3017 (2011). 36. Moraga, A. et al. Aging increases microglial proliferation, delays cell migration, and decreases cortical neurogenesis after focal cerebral ischemia. Journal of Neuroinflammation 12, 87 (2015). 105 37. Carmichael, S. T. Rodent models of focal stroke: size, mechanism, and purpose. NeuroRx   : the journal of the American Society for Experimental NeuroTherapeutics 2, 396–409 (2005). 38. García-Bonilla, L. et al. Recommendations guide for experimental animal models in stroke research. Neurología (English Edition) 26, 105–110 (2011). 39. Mergenthaler, P. & Meisel, a. Do stroke models model stroke? Disease Models & Mechanisms 5, 718–725 (2012). 40. Zhang, Z. et al. A new rat model of thrombotic focal cerebral ischemia. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 17, 123–135 (1997). 41. Orset, C. et al. Mouse model of in situ thromboembolic stroke and reperfusion. Stroke 38, 2771–2778 (2007). 42. Karatas, H. et al. Thrombotic distal middle cerebral artery occlusion produced by topical FeCl(3) application: a novel model suitable for intravital microscopy and thrombolysis studies. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 31, 1452–1460 (2011). 43. Morancho, A. et al. A new method for focal transient cerebral ischaemia by distal compression of the middle cerebral artery. Neuropathology and Applied Neurobiology 38, 617–627 (2012). 44. Lok, J. et al. Cell-cell signaling in the neurovascular unit. Neurochemical Research 32, 2032–2045 (2007). 45. Del Zoppo, G. J. Stroke and neurovascular protection. The New England journal of medicine 354, 553–555 (2006). 46. Abbott, N. J., Patabendige, A. A. K., Dolman, D. E. M., Yusof, S. R. & Begley, D. J. Structure and function of the blood-brain barrier. Neurobiology of Disease 37, 13–25 (2010). 47. Hawkins, B. T. & Davis, T. P. The Blood-Brain Barrier / Neurovascular Unit in Health and Disease. Pharmacological reviews 57, 173–185 (2005). 48. Sánchez-Mendoza, E. et al. Transient focal cerebral ischemia significantly alters not only EAATs but also VGLUTs expression in rats: relevance of changes in reactive astroglia. Journal of neurochemistry 113, 1343–55 (2010). 106 References 49. Takatsuru, Y. et al. Critical role of the astrocyte for functional remodeling in contralateral hemisphere of somatosensory cortex after stroke. The Journal of neuroscience   : the official journal of the Society for Neuroscience 33, 4683–92 (2013). 50. Murphy, T. H. & Corbett, D. Plasticity during stroke recovery: from synapse to behaviour. Nature reviews. Neuroscience 10, 861–872 (2009). 51. Lo, E. H. A new penumbra: transitioning from injury into repair after stroke. Nature medicine 14, 497–500 (2008). 52. Pascual-Leone, A., Amedi, A., Fregni, F. & Merabet, L. B. The plastic human brain cortex. Annual review of neuroscience 28, 377–401 (2005). 53. Manuscript, A. Experience, Cortical Remapping, and Recovery in Brain Disease. Neurobiol Dis 37, 997–1003 (2010). 54. Gage, F. H. Mammalian Neural Stem Cells. Science 287, 1433–1438 (2000). 55. Arvidsson, A., Collin, T., Kirik, D., Kokaia, Z. & Lindvall, O. Neuronal replacement from endogenous precursors in the adult brain after stroke. Nature medicine 8, 963–970 (2002). 56. Parent, J. M., Vexler, Z. S., Gong, C., Derugin, N. & Ferriero, D. M. Rat forebrain neurogenesis and striatal neuron replacement after focal stroke. Annals of Neurology 52, 802–813 (2002). 57. Marti-Fabregas, J. et al. Proliferation in the human ipsilateral subventricular zone after ischemic stroke. Neurology 74, 357–365 (2010). 58. Yamashita, T. et al. Subventricular zone-derived neuroblasts migrate and differentiate into mature neurons in the post-stroke adult striatum. The Journal of neuroscience  : the official journal of the Society for Neuroscience 26, 6627–6636 (2006). 59. Ohab, J. J., Fleming, S., Blesch, A. & Carmichael, S. T. A neurovascular niche for neurogenesis after stroke. The Journal of neuroscience : the official journal of the Society for Neuroscience 26, 13007–13016 (2006). 60. Thored, P. et al. Long-term neuroblast migration along blood vessels in an area with transient angiogenesis and increased vascularization after stroke. Stroke 38, 3032–3039 (2007). 61. Zhang, R. et al. Stroke transiently increases subventricular zone cell division from asymmetric to symmetric and increases neuronal differentiation in the 107 adult rat. The Journal of neuroscience : the official journal of the Society for Neuroscience 24, 5810–5815 (2004). 62. Pepper, M. S. Manipulating angiogenesis. From basic science to the bedside. [Review] [166 refs]. Arterioscler.Thromb.Vasc.Biol. 17, 605–619 (1997). 63. Asahara, T. et al. Isolation of putative progenitor endothelial cells for angiogenesis. Science (New York, N.Y.) 275, 964–967 (1997). 64. Vasa, M. et al. Number and migratory activity of circulating endothelial progenitor cells inversely correlate with risk factors for coronary artery disease. Circulation research 89, E1–E7 (2001). 65. Griese, D. P. et al. Isolation and Transplantation of Autologous Circulating Endothelial Cells into Denuded Vessels and Prosthetic Grafts: Implications for Cell-Based Vascular Therapy. Circulation 108, 2710–2715 (2003). 66. Hayashi, T., Noshita, N., Sugawara, T. & Chan, P. H. Temporal profile of angiogenesis and expression of related genes in the brain after ischemia. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 23, 166–180 (2003). 67. Schmidt, A., Brixius, K. & Bloch, W. Endothelial precursor cell migration during vasculogenesis. Circulation Research 101, 125–136 (2007). 68. Cui, L. et al. Stromal cell-derived factor-1 and its receptor CXCR4 in adult neurogenesis after cerebral ischemia. Restorative neurology and neuroscience 31, 239–51 (2013). 69. Page-McCaw, A., Serano, J., Santë, J. M. & Rubin, G. M. Drosophila matrix metalloproteinases are required for tissue remodeling, but not embryonic development. Developmental Cell 4, 95–106 (2003). 70. Porter, S., Clark, I. M., Kevorkian, L. & Edwards, D. R. The ADAMTS metalloproteinases. The Biochemical journal 386, 15–27 (2005). 71. Rosell, A. & Lo, E. H. Multiphasic roles for matrix metalloproteinases after stroke. Current Opinion in Pharmacology 8, 82–89 (2008). 72. Apte, S. S. & Parks, W. C. Metalloproteinases: A parade of functions in matrix biology and an outlook for the future. Matrix Biology 44-46, 1–6 (2015). 73. Klein, T. & Bischoff, R. Physiology and pathophysiology of matrix metalloproteases. Amino Acids 41, 271–290 (2011). 108 References 74. Fanjul-Fernández, M., Folgueras, A. R., Cabrera, S. & López-Otín, C. Matrix metalloproteinases: evolution, gene regulation and functional analysis in mouse models. Biochimica et biophysica acta 1803, 3–19 (2010). 75. Visse, R. & Nagase, H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: Structure, function, and biochemistry. Circulation Research 92, 827–839 (2003). 76. Clark, A. W., Krekoski, C. A., Bou, S. S., Chapman, K. R. & Edwards, D. R. Increased gelatinase A (MMP-2) and gelatinase B (MMP-9) activities in human brain after focal ischemia. Neuroscience Letters 238, 53–56 (1997). 77. Anthony, D. C. et al. Differential matrix metalloproteinase expression in cases of multiple sclerosis and stroke. Neuropathology and applied neurobiology 23, 406–15 (1997). 78. Heo, J. H. et al. Matrix metalloproteinases increase very early during experimental focal cerebral ischemia. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 19, 624–33 (1999). 79. Montaner, J. et al. Matrix metalloproteinase expression after human cardioembolic stroke: temporal profile and relation to neurological impairment. Stroke; a journal of cerebral circulation 32, (2001). 80. Alvarez-Sabín, J. et al. Temporal profile of matrix metalloproteinases and their inhibitors after spontaneous intracerebral hemorrhage: relationship to clinical and radiological outcome. Stroke; a journal of cerebral circulation 35, 1316–1322 (2004). 81. Cuadrado, E. et al. Vascular MMP-9/TIMP-2 and Neuronal MMP-10 Up-Regulation in Human Brain after Stroke: A Combined Laser Microdissection and Protein Array Study. Journal of Proteome Research 8, 3191–3197 (2008). 82. Rosell, a. et al. Increased Brain Expression of Matrix Metalloproteinase-9 After Ischemic and Hemorrhagic Human Stroke. Stroke 37, 1399–1406 (2006). 83. Rosell, A. et al. MMP-9-positive neutrophil infiltration is associated to blood-brain barrier breakdown and basal lamina type IV collagen degradation during hemorrhagic transformation after human ischemic stroke. Stroke 39, 1121–1126 (2008). 84. Asahi, M. et al. Role for matrix metalloproteinase 9 after focal cerebral ischemia: effects of gene knockout and enzyme inhibition with BB-94. Journal 109 of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 20, 1681–1689 (2000). 85. Lee, S.-R., Tsuji, K., Lee, S.-R. & Lo, E. H. Role of matrix metalloproteinases in delayed neuronal damage after transient global cerebral ischemia. The Journal of neuroscience : the official journal of the Society for Neuroscience 24, 671–8 (2004). 86. Pfefferkorn, T. & Rosenberg, G. a. Closure of the blood-brain barrier by matrix metalloproteinase inhibition reduces rtPA-mediated mortality in cerebral ischemia with delayed reperfusion. Stroke; a journal of cerebral circulation 34, 2025–30 (2003). 87. Lakhan, S. E., Kirchgessner, A., Tepper, D. & Leonard, A. Matrix metalloproteinases and blood-brain barrier disruption in acute ischemic stroke. Frontiers in Neurology 4 APR, 1–15 (2013). 88. Mun-Bryce, S. & Rosenberg, G. a. Matrix metalloproteinases in cerebrovascular disease. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 18, 1163–72 (1998). 89. Barkho, B. Z. et al. Endogenous matrix metalloproteinase (MMP)-3 and MMP-9 promote the differentiation and migration of adult neural progenitor cells in response to chemokines. Stem cells 26, 3139–3149 (2008). 90. Bergers, G. et al. Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nature cell biology 2, 737–44 (2000). 91. Tsai, P. T. et al. A critical role of erythropoietin receptor in neurogenesis and post-stroke recovery. The Journal of neuroscience : the official journal of the Society for Neuroscience 26, 1269–1274 (2006). 92. Lin, C.-Y. et al. Dynamic changes in vascular permeability, cerebral blood volume, vascular density, and size after transient focal cerebral ischemia in rats: evaluation with contrast-enhanced magnetic resonance imaging. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 28, 1491–501 (2008). 93. Zhao, B.-Q. et al. Role of matrix metalloproteinases in delayed cortical responses after stroke. Nature medicine 12, 441–445 (2006). 94. Lee, S.-R. et al. Involvement of matrix metalloproteinase in neuroblast cell migration from the subventricular zone after stroke. The Journal of 110 References neuroscience : the official journal of the Society for Neuroscience 26, 3491–3495 (2006). 95. Morancho, A. et al. Cerebral ischaemia and matrix metalloproteinase-9 modulate the angiogenic function of early and late outgrowth endothelial progenitor cells. Journal of Cellular and Molecular Medicine 17, 1543–1553 (2013). 96. Morancho, A. et al. Impaired vascular remodeling after endothelial progenitor cell transplantation in MMP9-deficient mice suffering cortical cerebral ischemia. Journal of Cerebral Blood Flow & Metabolism 1–5 (2015). doi:10.1038/jcbfm.2015.180 97. Balbín, M. et al. Expression and regulation of collagenase-3 (MMP-13) in human malignant tumors. APMIS : acta pathologica, microbiologica, et immunologica Scandinavica 107, 45–53 (1999). 98. Burrage, P. S., Mix, K. S. & Brinckerhoff, C. E. Matrix metalloproteinases: role in arthritis. Frontiers in bioscience : a journal and virtual library 11, 529–543 (2006). 99. Hattori, N. et al. MMP-13 plays a role in keratinocyte migration, angiogenesis, and contraction in mouse skin wound healing. The American journal of pathology 175, 533–546 (2009). 100. Nagel, S. et al. Focal cerebral ischemia induces changes in both MMP-13 and aggrecan around individual neurons. Brain Research 1056, 43–50 (2005). 101. Cuadrado, E. et al. Matrix metalloproteinase-13 is activated and is found in the nucleus of neural cells after cerebral ischemia. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 29, 398–410 (2009). 102. Rosell, A. et al. A matrix metalloproteinase protein array reveals a strong relation between MMP-9 and MMP-13 with diffusion-weighted image lesion increase in human stroke. Stroke 36, 1415–1420 (2005). 103. Shyam Prabhakaran, MD, MS; Ilana Ruff, MD; Richard A. Bernstein, MD, P. Acute Stroke Intervention. Current Problems in Cardiology 313, 1451–1462 (2015). 104. Bori I, M. C. Conceptos fundamentales en rehabilitación en el ictus:¿dónde, cómo, quién, cuándo y cuánta? Neuroreparación y Rehabilitación tras el ictus. (2010). 111 105. Langhorne, P., Bernhardt, J. & Kwakkel, G. Stroke rehabilitation. Lancet 377, 1693–702 (2011). 106. Langhorne, P. et al. What are the components of effective stroke unit care? Age and Ageing 31, 365–371 (2002). 107. Teasell, R. et al. The next revolution in stroke care. Expert Review of Neurotherapeutics 14, 1307–1314 (2014). 108. Brewer, L., Horgan, F., Hickey, a & Williams, D. Stroke rehabilitation: recent advances and future therapies. QJM : monthly journal of the Association of Physicians 106, 11–25 (2013). 109. Langhammer, B., Lindmark, B. & Stanghelle, J. K. Stroke patients and long-term training: is it worthwhile? A randomized comparison of two different training strategies after rehabilitation. Clinical rehabilitation 21, 495–510 (2007). 110. Langhammer, B. & Lindmark, B. Functional exercise and physical fitness post stroke: the importance of exercise maintenance for motor control and physical fitness after stroke. Stroke research and treatment 2012, 864835 (2012). 111. Pekna, M., Pekny, M. & Nilsson, M. Modulation of Neural Plasticity as a Basis for Stroke Rehabilitation. Stroke 43, 2819–2828 (2012). 112. Rosell, A. et al. Factors Secreted by Endothelial Progenitor Cells Enhance Neurorepair Responses after Cerebral Ischemia in Mice. PLoS ONE 8, 1–10 (2013). 113. Carmichael, S. T. Cellular and molecular mechanisms of neural repair after stroke: Making waves. Annals of Neurology 59, 735–742 (2006). 114. Manolescu, B. N. et al. Dynamics of inflammatory markers in post-acute stroke patients undergoing rehabilitation. Inflammation 34, 551–8 (2011). 115. Blicher, J. U. et al. GABA Levels Are Decreased After Stroke and GABA Changes During Rehabilitation Correlate With Motor Improvement. Neurorehabilitation and neural repair 29, 278–86 (2015). 116. Hsieh, Y. W. et al. The reliability and predictive ability of a biomarker of oxidative DNA damage on functional outcomes after stroke rehabilitation. International Journal of Molecular Sciences 15, 6504–6516 (2014). 112 References 117. Chopp, J. Z. and M. Cell-based therapy for ischemic stroke. 13, 1229–1240 (2013). 118. Kalladka, D. & Muir, K. W. Brain repair: Cell therapy in stroke. Stem Cells and Cloning: Advances and Applications 7, 31–44 (2014). 119. Savitz, S. I., Cramer, S. C. & Wechsler, L. Stem cells as an emerging paradigm in stroke 3 enhancing the development of clinical trials. Stroke 45, 634–639 (2014). 120. Asahara, T. et al. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circulation research 85, 221–228 (1999). 121. Urbich, C. & Dimmeler, S. Endothelial progenitor cells: Characterization and role in vascular biology. Circulation Research 95, 343–353 (2004). 122. Navarro-Sobrino, M. et al. Mobilization, endothelial differentiation and functional capacity of endothelial progenitor cells after ischemic stroke. Microvascular Research 80, 317–323 (2010). 123. Park, K.-P. et al. Plasma and brain matrix metalloproteinase-9 after acute focal cerebral ischemia in rats. Stroke; a journal of cerebral circulation 40, 2836–42 (2009). 124. Morancho, A., Rosell, A., García-Bonilla, L. & Montaner, J. Metalloproteinase and stroke infarct size: Role for anti-inflammatory treatment? Annals of the New York Academy of Sciences 1207, 123–133 (2010). 125. Cuadrado, E. et al. Vascular MMP-9/TIMP-2 and Neuronal MMP-10 Up-Regulation in Human Brain after Stroke: A Combined Laser Microdissection and Protein Array Study. Journal of Proteome Research 8, 3191–3197 (2009). 126. Morancho, A. et al. Impaired vascular remodeling after endothelial progenitor cell transplantation in MMP9-deficient mice suffering cortical cerebral ischemia. Journal of Cerebral Blood Flow & Metabolism 1–5 (2015). doi:10.1038/jcbfm.2015.180 127. Lee, C. Z., Xue, Z., Zhu, Y., Yang, G.-Y. & Young, W. L. Matrix metalloproteinase-9 inhibition attenuates vascular endothelial growth factor-induced intracerebral hemorrhage. Stroke; a journal of cerebral circulation 38, 2563–2568 (2007). 128. Duncan, M. E., Richardson, J. P., Murray, G. I., Melvin, W. T. & Fothergill, J. E. Human matrix metalloproteinase-9: activation by limited trypsin treatment 113 and generation of monoclonal antibodies specific for the activated form. European journal of biochemistry / FEBS 258, 37–43 (1998). 129. Lenglet, S. et al. Analysis of the expression of nine secreted matrix metalloproteinases and their endogenous inhibitors in the brain of mice subjected to ischaemic stroke. Thrombosis and Haemostasis 112, 363–378 (2014). 130. Rosell, A. et al. Distal Occlusion of the Middle Cerebral Artery in Mice: Are We Ready to Assess Long-Term Functional Outcome? Translational Stroke Research 4, 297–307 (2013). 131. Zarruk, J. G. et al. Neurological tests for functional outcome assessment in rodent models of ischaemic stroke. Rev Neurol 53, 607–618 (2011). 132. Shelton, S. B. et al. A simple, efficient tool for assessment of mice after unilateral cortex injury. Journal of Neuroscience Methods 168, 431–442 (2008). 133. Bouët, V. et al. Sensorimotor and cognitive deficits after transient middle cerebral artery occlusion in the mouse. Experimental Neurology 203, 555–567 (2007). 134. Zhang, L., Chen, J., Li, Y., Zhang, Z. G. & Chopp, M. Quantitative measurement of motor and somatosensory impairments after mild (30 min) and severe (2 h) transient middle cerebral artery occlusion in rats. Journal of the Neurological Sciences 174, 141–146 (2000). 135. Brott, T. et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke; a journal of cerebral circulation 20, 864–870 (1989). 136. Van Swieten, J. C., Koudstaal, P. J., Visser, M. C., Schouten, H. J. & van Gijn, J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke; a journal of cerebral circulation 19, 604–607 (1988). 137. Saver, J. L. Optimal end points for acute stroke therapy trials: best ways to measure treatment effects of drugs and devices. Stroke; a journal of cerebral circulation 42, 2356–62 (2011). 138. Tan, S. et al. Frequency and predictors of spontaneous hemorrhagic transformation in ischemic stroke and its association with prognosis. Journal of Neurology 261, 905–912 (2014). 114 References 139. Latour, L. L., Kang, D.-W., Ezzeddine, M. A., Chalela, J. A. & Warach, S. Early blood-brain barrier disruption in human focal brain ischemia. Annals of neurology 56, 468–477 (2004). 140. Wang, X. et al. Lipoprotein receptor-mediated induction of matrix metalloproteinase by tissue plasminogen activator. Nature medicine 9, 1313–1317 (2003). 141. Del Zoppo, G. J. et al. Microglial cell activation is a source of metalloproteinase generation during hemorrhagic transformation. Journal of Cerebral Blood Flow & Metabolism 32, 919–932 (2012). 142. Shimazawa, M. & Hara, H. Novel situations of endothelial injury in stroke--mechanisms of stroke and strategy of drug development: Protective effects of antiplatelet agents against stroke. Journal of pharmacological sciences 116, 30–35 (2011). 143. Yang, Y. & Rosenberg, G. a. Blood-brain barrier breakdown in acute and chronic cerebrovascular disease. Stroke 42, 3323–3328 (2011). 144. Wang, C. X. & Shuaib, A. Critical role of microvasculature basal lamina in ischemic brain injury. Progress in neurobiology 83, 140–148 (2007). 145. Sandoval, K. E. & Witt, K. A. Blood-brain barrier tight junction permeability and ischemic stroke. Neurobiology of Disease 32, 200–219 (2008). 146. Kanazawa, M. et al. Inhibition of VEGF signaling pathway attenuates hemorrhage after tPA treatment. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 31, 1461–1474 (2011). 147. Terruso, V. et al. Frequency and determinants for hemorrhagic transformation of cerebral infarction. Neuroepidemiology 33, 261–265 (2009). 148. Zhu, Y. et al. Angiopoietin-2 facilitates vascular endothelial growth factor-induced angiogenesis in the mature mouse brain. Stroke 36, 1533–1537 (2005). 149. Zechariah, A. et al. Vascular endothelial growth factor promotes pericyte coverage of brain capillaries, improves cerebral blood flow during subsequent focal cerebral ischemia, and preserves the metabolic penumbra. Stroke 44, 1690–1697 (2013). 150. Kilic, E. et al. Delayed melatonin administration promotes neuronal survival, neurogenesis and motor recovery, and attenuates hyperactivity and anxiety 115 after mild focal cerebral ischemia in mice. Journal of Pineal Research 45, 142–148 (2008). 151. Ueno, M. et al. The expression of matrix metalloproteinase-13 is increased in vessels with blood–brain barrier impairment in a stroke-prone hypertensive model. Hypertension Research 32, 332–338 (2009). 152. Hadler-Olsen, E., Fadnes, B., Sylte, I., Uhlin-Hansen, L. & Winberg, J. O. Regulation of matrix metalloproteinase activity in health and disease. FEBS Journal 278, 28–45 (2011). 153. Itoh, Y. Membrane-type matrix metalloproteinases: Their functions and regulations. Matrix Biology 44-46, 207–223 (2015). 154. Chelluboina, B. et al. Post-transcriptional inactivation of matrix metalloproteinase-12 after focal cerebral ischemia attenuates brain damage. Scientific Reports 5, 9504 (2015). 155. Ma, F., Morancho, A., Montaner, J. & Rosell, A. Endothelial progenitor cells and revascularization following stroke. Brain Research 1623, 150–159 (2015). 156. Moraga, A. et al. Toll-like receptor 4 modulates cell migration and cortical neurogenesis after focal cerebral ischemia. FASEB journal 1–9 (2014). doi:10.1096/fj.14-252452 157. Saha, B., Peron, S., Murray, K., Jaber, M. & Gaillard, A. Cortical lesion stimulates adult subventricular zone neural progenitor cell proliferation and migration to the site of injury. Stem Cell Research 11, 965–977 (2013). 158. Ohab, J. J. & Carmichael, S. T. Poststroke neurogenesis: emerging principles of migration and localization of immature neurons. The Neuroscientist : a review journal bringing neurobiology, neurology and psychiatry 14, 369–380 (2008). 159. Thored, P. et al. Persistent production of neurons from adult brain stem cells during recovery after stroke. Stem cells 24, 739–747 (2006). 160. Kahle, M. P. & Bix, G. J. Neuronal restoration following ischemic stroke: influences, barriers, and therapeutic potential. Neurorehabilitation and neural repair 27, 469–78 (2013). 161. Licht, T. & Keshet, E. Delineating multiple functions of VEGF-A in the adult brain. Cellular and Molecular Life Sciences 70, 1727–1737 (2013). 162. Greenberg, D. A. & Jin, K. Vascular endothelial growth factors (VEGFs) and stroke. Cellular and molecular life sciences : CMLS 70, 1753–61 (2013). 116 References 163. Kovács, Z., Ikezaki, K., Samoto, K., Inamura, T. & Fukui, M. VEGF and flt. Expression time kinetics in rat brain infarct. Stroke; a journal of cerebral circulation 27, 1865–1872; discussion 1872–1873 (1996). 164. Sun, Y. et al. VEGF-induced neuroprotection, neurogenesis, and angiogenesis after focal cerebral ischemia. Journal of Clinical Investigation 111, 1843–1851 (2003). 165. Hareng, L. & Hartung, T. Induction and regulation of endogenous granulocyte colony-stimulating factor formation. Biological Chemistry 383, 1501–1517 (2002). 166. Schneider, A. et al. The hematopoietic factor G-CSF is a neuronal ligand that counteracts programmed cell death and drives neurogenesis. Journal of Clinical Investigation 115, 2083–2098 (2005). 167. Hasselblatt, M., Jeibmann, A., Riesmeier, B., Maintz, D. & Schäbitz, W. R. Granulocyte-colony stimulating factor (G-CSF) and G-CSF receptor expression in human ischemic stroke. Acta Neuropathologica 113, 45–51 (2007). 168. Sehara, Y. et al. Potentiation of neurogenesis and angiogenesis by G-CSF after focal cerebral ischemia in rats. Brain Research 1151, 142–149 (2007). 169. Maisonpierre, P. C. et al. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science (New York, N.Y.) 277, 55–60 (1997). 170. Wang, R.-G. & Zhu, X.-Z. Expression of angiopoietin-2 and vascular endothelial growth factor in mice cerebral cortex after permanent focal cerebral ischemia. Acta pharmacologica Sinica 23, 405–411 (2002). 171. Liu, X. S. et al. Angiopoietin 2 mediates the differentiation and migration of neural progenitor cells in the subventricular zone after stroke. Journal of Biological Chemistry 284, 22680–22689 (2009). 172. Nageswari, K., Mizusawa, S., Kondoh, Y., Nakamura, K. & Kanno, I. Therapeutic efficacy of basic fibroblast growth factor on experimental focal ischemia studied by magnetic resonance imaging. Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association 14, 187–192 (2005). 173. Cheng, X. et al. Acidic fibroblast growth factor delivered intranasally induces neurogenesis and angiogenesis in rats after ischemic stroke. Neurological research 33, 675–80 (2011). 117 174. Iihara, K., Sasahara, M., Hashimoto, N. & Hazama, F. Induction of platelet-derived growth factor beta-receptor in focal ischemia of rat brain. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 16, 941–949 (1996). 175. Krupinski, J. et al. A Putative Role for Platelet-Derived Growth Factor in Angiogenesis and Neuroprotection After Ischemic Stroke in Humans. Stroke 28, 564–573 (1997). 176. Hayon, Y., Dashevsky, O., Shai, E., Varon, D. & Leker, R. R. Platelet microparticles promote neural stem cell proliferation, survival and differentiation. Journal of molecular neuroscience : MN 47, 659–65 (2012). 177. Takahashi, T. et al. Ischemia- and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nature medicine 5, 434–438 (1999). 178. Carmeliet, P. Angiogenesis in health and disease. Nature medicine 9, 653–660 (2003). 179. Taguchi, A. et al. Circulating CD34-positive cells provide an index of cerebrovascular function. Circulation 109, 2972–2975 (2004). 180. Mao, L. et al. Endogenous endothelial progenitor cells participate in neovascularization via CXCR4/SDF-1 axis and improve outcome after stroke. CNS Neuroscience and Therapeutics 20, 460–468 (2014). 181. Massot, a. et al. Decreased levels of angiogenic growth factors in intracranial atherosclerotic disease despite severity-related increase in endothelial progenitor cell counts. Cerebrovascular Diseases 35, 81–88 (2013). 182. Urbich, C. et al. Proteomic characterization of human early pro-angiogenic cells. Journal of molecular and cellular cardiology 50, 333–6 (2011). 183. Navarro-Sobrino, M. et al. The angiogenic gene profile of circulating endothelial progenitor cells from ischemic stroke patients. Vascular cell 5, 3 (2013). 184. Orlic, D. et al. Bone marrow cells regenerate infarcted myocardium. Nature 410, 701–705 (2001). 185. Zhang, Z. G., Zhang, L., Jiang, Q. & Chopp, M. Bone marrow-derived endothelial progenitor cells participate in cerebral neovascularization after 118 References focal cerebral ischemia in the adult mouse. Circulation Research 90, 284–288 (2002). 186. Rehman, J. Peripheral Blood ‘Endothelial Progenitor Cells’ Are Derived From Monocyte/Macrophages and Secrete Angiogenic Growth Factors. Circulation 107, 1164–1169 (2003). 187. Yang, Z. et al. Paracrine factors secreted by endothelial progenitor cells prevent oxidative stress-induced apoptosis of mature endothelial cells. Atherosclerosis 211, 103–109 (2010). 188. Leeman, M. F., Curran, S. & Murray, G. I. The structure, regulation, and function of human matrix metalloproteinase-13. Critical reviews in biochemistry and molecular biology 37, 149–166 (2002). 189. Huang, P. H. et al. Matrix metalloproteinase-9 is essential for ischemia-induced neovascularization by modulating bone marrow-derived endothelial progenitor Cells. Arteriosclerosis, Thrombosis, and Vascular Biology 29, 1179–1184 (2009). 190. Sieveking, D. P., Buckle, A., Celermajer, D. S. & Ng, M. K. C. Strikingly Different Angiogenic Properties of Endothelial Progenitor Cell Subpopulations. Insights From a Novel Human Angiogenesis Assay. Journal of the American College of Cardiology 51, 660–668 (2008). 191. Burlacu, A., Grigorescu, G., Rosca, A.-M., Preda, M. B. & Simionescu, M. Factors Secreted by Mesenchymal Stem Cells and Endothelial Progenitor Cells Have Complementary Effects on Angiogenesis In Vitro. Stem Cells and Development 00, 121019095328006 (2012). 192. Cucchiara, B. & Montaner, J. Blood biomarkers for predicting outcome after stroke: Reading the tea leaves. Neurology 80, 1270–1271 (2013). 193. Hasan, N., McColgan, P., Bentley, P., Edwards, R. J. & Sharma, P. Towards the identification of blood biomarkers for acute stroke in humans: A comprehensive systematic review. British Journal of Clinical Pharmacology 74, 230–240 (2012). 194. Montaner, J. et al. Plasmatic level of neuroinflammatory markers predict the extent of diffusion-weighted image lesions in hyperacute stroke. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 23, 1403–1407 (2003). 195. Ning, M. et al. Association between tPA therapy and raised early matrix metalloproteinase-9 in acute stroke. Neurology 66, 1550–1555 (2006). 119 196. Castellanos, M. et al. Plasma metalloproteinase-9 concentration predicts hemorrhagic transformation in acute ischemic stroke. Stroke 34, 40–45 (2003). 197. Jickling, G. C. & Sharp, F. R. Biomarker Panels in Ischemic Stroke. Stroke 915–920 (2015). doi:10.1161/STROKEAHA.114.005604 198. McCawley, L. J. & Matrisian, L. M. Matrix metalloproteinases: They’re not just for matrix anymore! Current Opinion in Cell Biology 13, 534–540 (2001). 199. Van Hove, I., Lemmens, K., Van De Velde, S., Verslegers, M. & Moons, L. Matrix metalloproteinase-3 in the central nervous system: A look on the bright side. Journal of Neurochemistry 123, 203–216 (2012). 200. Van Hove, I. et al. An aberrant cerebellar development in mice lacking matrix metalloproteinase-3. Molecular neurobiology 45, 17–29 (2012). 201. Suzuki, Y., Nagai, N., Umemura, K., Collen, D. & Lijnen, H. R. Stromelysin-1 (MMP-3) is critical for intracranial bleeding after t-PA treatment of stroke in mice. Journal of Thrombosis and Haemostasis 5, 1732–1739 (2007). 202. Solé, S., Petegnief, V., Gorina, R., Chamorro, A. & Planas, A. M. Activation of matrix metalloproteinase-3 and agrin cleavage in cerebral ischemia/reperfusion. Journal of neuropathology and experimental neurology 63, 338–49 (2004). 203. Yang, Y. et al. Early inhibition of MMP activity in ischemic rat brain promotes expression of tight junction proteins and angiogenesis during recovery. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 33, 1104–14 (2013). 204. P.W., D. Outcome measures in stroke rehabilitation. Handbook of Clinical Neurology 110, 105–111 (2013). 205. Valavanidis, A., Vlachogianni, T. & Fiotakis, C. 8-hydroxy-2’ -deoxyguanosine (8-OHdG): A critical biomarker of oxidative stress and carcinogenesis. Journal of environmental science and health. Part C, Environmental carcinogenesis & ecotoxicology reviews 27, 120–139 (2009). 120 Supplementary Supplementary Information 1 Curriculum vitae of the author 121 122 Curriculum Vitae Personal information: Name: Feifei Surname: Ma Date of birth: 11/08/1987 Academic records: 2012 November - 2015 November Doctor degree candidate, major in Neuroscience, Universitat Autonoma de Barcelona, Vall d’Hebron Institut de Recerca, with the project of “Neuroprotection and neurorepair treatments for ischemic stroke” 2009 September - 2012 July Master degree, major in Neuropharmacology, Zhengzhou University, with the project of “Neuroprotective potential and mechanism of 3-butyl-6-bromol (3H)-isobenzofuranone on cerebral ischemia in rats” Relevant skills and technical knowledge: Middle cerebral artery occlusion (MCAo) model in mice and rats Animal experimental techniques Immunology experimental skills (Immunohistochemistry, Immunofluorescence, Luminex assay, ELISA) RT-PCR and Western blot techniques Use of confocal microscope, Image J software, platelet aggregation instrument Primary endothelial progenitor cells obtain Statistical analyses with SPSS software Manuscripts under submission 1. F. Ma, S. Rodriguez, X. Buxo, A. Morancho, I. Riba-Llena, A. Crrera, A. Bustamante, D. Giralt, J. Montaner, C. Martinez, I. Bori and A. Rosell Plasma matrix metalloproteinase levels in stroke patients during intensive rehabilitation therapy. In preparation. 123 2. F. Ma, P. Martínez-San Segundo, V. Barceló, A. Morancho, M. Gabriel-Salazar, D. Giralt, J. Montaner and A. Rosell Matrix Metalloproteinase-13 Controls Neuroprotection and Neurorepair after Cerebral Cortical Ischemia in Mice. Neurobiology of Disease, under review. Publications 1. A. Morancho, F. Ma, V. Barceló, D. Giralt, J. Montaner and A. Rosell Impaired vascular remodeling after endothelial progenitor cell transplantation in MMP9-deficient mice suffering cortical cerebral ischemia. Journal of Cerebral Blood Flow & Metabolism (2015), 1-5. 2. F. Ma, A. Morancho, J. Montaner and A. Rosell Endothelial progenitor cells and revascularization following stroke. Brain Research. 1623 (2015) 150-159. 3. F. Ma, Y. Gao, H. Qiao, X. Hu, J. Chang Antiplatelet activity of 3-butyl-6-bromo -1(3H)-isobenzofuranone on rat platelet aggregation. Journal of Thromb and Thrombolysis (2012) 33:64-73. Presentation and poster 1. F. Ma, A. Morancho, P. Martínez-San Segundo, M. Gabriel-Salazar, V. Barceló, D. Giralt, J. Montaner and A. Rosell MATIX METALLOPROTEINASE-13 (MMP-13) MODULATES NEUROPROTECTION AND NEUROREPAIR AFTER CEREBRAL CORTICAL ISCHEMIA. 1st European Stroke Organization conference (April, 2015) 2. F. Ma, V. Barceló, A. Morancho, P. Martínez-San Segundo, M. Gabriel-Salazar, D. Giralt, J. Montaner and A. Rosell Matrix metalloproteinase (MMP)-13 modulates neuroprotection and neuro-repair after cerebral cortical ischemia. 8th VHIR Scientific Session (December 2014) 124 Supplementary Supplementary Information 2 Additional publication and poster 125 126 127 128 129 130 131 132 133 134